{
  "schema_version": "2026-05-07",
  "generated_at": "2026-05-10T15:36:18.353Z",
  "source": "yangmao.ai provider YAML database",
  "license": "For citation and research use. Link back to https://yangmao.ai/en/tools/ai-free-api-checker/",
  "total_providers": 90,
  "free_api_providers": 44,
  "free_tier_providers": 87,
  "providers": [
    {
      "id": "yi",
      "name": "01.AI (Yi)",
      "name_zh": "零一万物",
      "url": "https://www.lingyiwanwu.com",
      "country": "cn",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/yi/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "¥10",
      "api_rate_limit": "5 RPM",
      "api_notes": "¥10 free credits on signup",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "By 01.AI, direct access in China. Yi-Lightning is very cost-effective.",
      "models": [
        {
          "name": "Yi-Lightning",
          "context": "16k",
          "rate_limit": "Fair use",
          "notes": "Free via Wanzhi"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "chat"
      ],
      "proof_url": "https://www.lingyiwanwu.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "claude-api",
      "name": "Anthropic API",
      "name_zh": "Anthropic API",
      "url": "https://console.anthropic.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/claude-api/",
      "has_free_tier": false,
      "free_tier_limit": "N/A",
      "has_free_api": true,
      "api_free_credits": "$5",
      "api_rate_limit": "5 RPM",
      "api_notes": "$5 free credits on signup",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Stable API, moderate latency.",
      "models": [
        {
          "name": "Claude Opus 4",
          "context": "200k",
          "rate_limit": "Paid only",
          "notes": "Strongest model, paid only"
        }
      ],
      "tags": [
        "api",
        "llm",
        "coding"
      ],
      "proof_url": "https://console.anthropic.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "anyscale",
      "name": "Anyscale",
      "name_zh": "Anyscale",
      "url": "https://anyscale.com",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/anyscale/",
      "has_free_tier": true,
      "free_tier_limit": "credit-based",
      "has_free_api": true,
      "api_free_credits": "$10 free credits",
      "api_rate_limit": "30 RPM",
      "api_notes": "New users get $10 free credits, built by the Ray framework team",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Use openllmapi.com as proxy.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "Available in free tier, accelerated by Ray framework"
        },
        {
          "name": "Mixtral 8x22B",
          "context": "64k",
          "rate_limit": "30 RPM",
          "notes": "Large MoE model"
        }
      ],
      "tags": [
        "api",
        "chat",
        "coding"
      ],
      "proof_url": "https://anyscale.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "baichuan",
      "name": "Baichuan AI",
      "name_zh": "百川智能",
      "url": "https://www.baichuan-ai.com",
      "country": "cn",
      "categories": [
        "chat",
        "reasoning"
      ],
      "provider_page": "/en/providers/baichuan/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "500万 tokens",
      "api_rate_limit": "5 RPM",
      "api_notes": "5M tokens free on signup",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese platform, direct access.",
      "models": [
        {
          "name": "Baichuan 4",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Baichuan's latest flagship model"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "chat",
        "api"
      ],
      "proof_url": "https://www.baichuan-ai.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cerebras",
      "name": "Cerebras",
      "name_zh": "Cerebras",
      "url": "https://cloud.cerebras.ai",
      "country": "us",
      "categories": [
        "api",
        "chat"
      ],
      "provider_page": "/en/providers/cerebras/",
      "has_free_tier": true,
      "free_tier_limit": "1M tokens/day",
      "has_free_api": true,
      "api_free_credits": "1M tokens/day",
      "api_rate_limit": "30 RPM / 60K TPM / 1M TPD",
      "api_notes": "No credit card, 1M tokens/day, OpenAI-compatible",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Extremely fast, ideal for low-latency use cases.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128K",
          "rate_limit": "30 RPM / 60K TPM",
          "notes": "World's fastest inference, 2000+ tokens/s"
        },
        {
          "name": "Llama 3.1 8B",
          "context": "128K",
          "rate_limit": "30 RPM / 60K TPM",
          "notes": "Lightweight and fast"
        }
      ],
      "tags": [
        "api",
        "fast-inference",
        "llm",
        "free"
      ],
      "proof_url": "https://cloud.cerebras.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "zhipu",
      "name": "ChatGLM (Zhipu AI)",
      "name_zh": "智谱清言 (智谱AI)",
      "url": "https://chatglm.cn",
      "country": "cn",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/zhipu/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "500万 tokens",
      "api_rate_limit": "5 RPM",
      "api_notes": "5M tokens free on signup",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese platform, direct access. GLM-4-Flash is free.",
      "models": [
        {
          "name": "GLM-4",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Zhipu's flagship model, free to use"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "chat",
        "api"
      ],
      "proof_url": "https://chatglm.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cloudflare-workers-ai",
      "name": "Cloudflare Workers AI",
      "name_zh": "Cloudflare Workers AI",
      "url": "https://ai.cloudflare.com",
      "country": "us",
      "categories": [
        "api",
        "chat",
        "image",
        "embedding"
      ],
      "provider_page": "/en/providers/cloudflare-workers-ai/",
      "has_free_tier": true,
      "free_tier_limit": "10,000 free requests/day",
      "has_free_api": true,
      "api_free_credits": "每天 10000 神经元（永久有效）",
      "api_rate_limit": "10000 requests/day",
      "api_notes": "Cloudflare Workers $5/mo plan includes Workers AI with 10,000 free neurons/day.\n50+ open-source models available including LLM, image generation (SD XL), embeddings, speech-to-text.\nOpenAI-compatible API via AI Gateway, or direct Workers AI binding.\nInference runs on Cloudflare's global edge network with ultra-low latency. Direct China access.\n",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Direct access from China via Cloudflare edge network, low latency. Workers AI accelerated by global CDN.",
      "models": [
        {
          "name": "@cf/meta/llama-3.1-8b-instruct",
          "context": "128k",
          "rate_limit": "10000/day",
          "notes": "Meta Llama 3.1 8B, lightweight chat model"
        },
        {
          "name": "@cf/meta/llama-3.3-70b-instruct-fp8-fast",
          "context": "128k",
          "rate_limit": "10000/day",
          "notes": "Llama 3.3 70B FP8 accelerated"
        },
        {
          "name": "@cf/google/gemma-7b-it-lora",
          "context": "8k",
          "rate_limit": "10000/day",
          "notes": "Google Gemma 7B with LoRA fine-tuning support"
        },
        {
          "name": "@cf/stabilityai/stable-diffusion-xl-base-1.0",
          "context": "",
          "rate_limit": "10000/day",
          "notes": "Stable Diffusion XL image generation, completely free"
        },
        {
          "name": "@cf/baai/bge-base-en-v1.5",
          "context": "",
          "rate_limit": "10000/day",
          "notes": "BGE embedding model for RAG and semantic search"
        },
        {
          "name": "@cf/microsoft/phi-2",
          "context": "2k",
          "rate_limit": "10000/day",
          "notes": "Microsoft Phi-2 small model"
        },
        {
          "name": "@cf/mistral/mistral-7b-instruct-v0.2-lora",
          "context": "32k",
          "rate_limit": "10000/day",
          "notes": "Mistral 7B with LoRA support"
        }
      ],
      "tags": [
        "api",
        "edge",
        "image",
        "embedding",
        "free",
        "cloudflare"
      ],
      "proof_url": "https://ai.cloudflare.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cohere",
      "name": "Cohere",
      "name_zh": "Cohere",
      "url": "https://cohere.com",
      "country": "us",
      "categories": [
        "api",
        "chat"
      ],
      "provider_page": "/en/providers/cohere/",
      "has_free_tier": true,
      "free_tier_limit": "1,000 calls/month (Trial Key)",
      "has_free_api": true,
      "api_free_credits": "1000 calls/month",
      "api_rate_limit": "Trial rate limits",
      "api_notes": "Free Trial Key, 1,000 calls/month across all models, no credit card, resets monthly",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy access.",
      "models": [
        {
          "name": "Command R+",
          "context": "128K",
          "rate_limit": "Trial 限速",
          "notes": "Top RAG and chat model"
        },
        {
          "name": "Rerank",
          "context": "N/A",
          "rate_limit": "Trial 限速",
          "notes": "Document reranking for RAG"
        },
        {
          "name": "Embed",
          "context": "N/A",
          "rate_limit": "Trial 限速",
          "notes": "Multilingual text embeddings"
        }
      ],
      "tags": [
        "api",
        "rag",
        "embeddings",
        "rerank",
        "free"
      ],
      "proof_url": "https://cohere.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "coze",
      "name": "Coze (ByteDance)",
      "name_zh": "扣子 (字节跳动)",
      "url": "https://www.coze.cn",
      "country": "cn",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/coze/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "Varies",
      "api_notes": "Free Bot API, build custom agents",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "By ByteDance. China version coze.cn direct access. International coze.com needs proxy.",
      "models": [
        {
          "name": "多模型可选",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Access to Doubao, GPT-4o and other models"
        }
      ],
      "tags": [
        "platform",
        "agents",
        "chat",
        "api"
      ],
      "proof_url": "https://www.coze.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "deepseek",
      "name": "DeepSeek",
      "name_zh": "DeepSeek",
      "url": "https://platform.deepseek.com",
      "country": "cn",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/deepseek/",
      "has_free_tier": true,
      "free_tier_limit": "50 requests/day",
      "has_free_api": true,
      "api_free_credits": "$5",
      "api_rate_limit": "2 RPM",
      "api_notes": "$5 free credits on signup, pay-as-you-go after. V4-Pro 75% off extended to 2026/05/31, cache hit price reduced to 1/10. V4-Flash $0.14/M input, $0.28/M output",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese service, direct access in China, very fast, no proxy needed",
      "models": [
        {
          "name": "DeepSeek-V4-Pro",
          "context": "1000k",
          "rate_limit": "Rate limited",
          "notes": "Released April 2026, 1.6T param MoE (49B active), 1M context, thinking/non-thinking modes"
        },
        {
          "name": "DeepSeek-V4-Flash",
          "context": "1000k",
          "rate_limit": "Rate limited",
          "notes": "Lightweight V4, 284B params (13B active), 1M context, excellent value"
        },
        {
          "name": "DeepSeek-V3",
          "context": "64k",
          "rate_limit": "50 RPD",
          "notes": "Previous flagship, still available"
        },
        {
          "name": "DeepSeek-R1",
          "context": "64k",
          "rate_limit": "50 RPD",
          "notes": "Reasoning model, strong at math/code"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "api",
        "chat",
        "coding"
      ],
      "proof_url": "https://platform.deepseek.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "lepton",
      "name": "DGX Cloud Lepton (formerly Lepton AI)",
      "name_zh": "DGX Cloud Lepton (原 Lepton AI)",
      "url": "https://build.nvidia.com/explore/discover",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/lepton/",
      "has_free_tier": true,
      "free_tier_limit": "10M tokens/day",
      "has_free_api": true,
      "api_free_credits": "$10 free credits",
      "api_rate_limit": "10 RPM",
      "api_notes": "New users get $10 free credits, founded by Yangqing Jia (PyTorch co-creator)",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Founded by Chinese-American team, good China access. API is directly accessible.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "10 RPM",
          "notes": "Free to use, built by Chinese-American team"
        },
        {
          "name": "Mixtral 8x7B",
          "context": "32k",
          "rate_limit": "10 RPM",
          "notes": "Available in free tier"
        }
      ],
      "tags": [
        "api",
        "chat",
        "coding"
      ],
      "proof_url": "https://build.nvidia.com/explore/discover",
      "last_verified": "2026-05-10"
    },
    {
      "id": "doubao",
      "name": "Doubao (ByteDance)",
      "name_zh": "豆包 (字节跳动)",
      "url": "https://www.doubao.com",
      "country": "cn",
      "categories": [
        "chat",
        "image",
        "video",
        "coding"
      ],
      "provider_page": "/en/providers/doubao/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "50万 tokens",
      "api_rate_limit": "5 RPM",
      "api_notes": "Free credits via Volcengine. Seed 2.0 Mini at ¥0.3/M tokens is almost free.",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese platform by ByteDance, direct access, fast. Servers across China.",
      "models": [
        {
          "name": "Seed 2.0 Pro",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Flagship all-rounder, close to DeepSeek V3"
        },
        {
          "name": "Seed 2.0 Code",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Coding-focused, surpasses DeepSeek V3 in coding"
        },
        {
          "name": "Seed 2.0 Mini",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Ultra-fast and cheap, ¥0.3/M tokens, almost free"
        },
        {
          "name": "Seed 2.0 Lite",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Balanced version, good value"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "image",
        "video",
        "coding",
        "cheap"
      ],
      "proof_url": "https://www.doubao.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "elevenlabs",
      "name": "ElevenLabs",
      "name_zh": "ElevenLabs",
      "url": "https://elevenlabs.io",
      "country": "us",
      "categories": [
        "audio"
      ],
      "provider_page": "/en/providers/elevenlabs/",
      "has_free_tier": true,
      "free_tier_limit": "10,000 characters/month",
      "has_free_api": true,
      "api_free_credits": "10K chars/month",
      "api_rate_limit": "Varies",
      "api_notes": "Free API: 10K characters per month",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Industry-leading TTS quality.",
      "models": [
        {
          "name": "ElevenLabs TTS",
          "context": "N/A",
          "rate_limit": "10K chars/month",
          "notes": "Most realistic AI text-to-speech"
        }
      ],
      "tags": [
        "tts",
        "audio",
        "voice",
        "api"
      ],
      "proof_url": "https://elevenlabs.io",
      "last_verified": "2026-05-10"
    },
    {
      "id": "ernie",
      "name": "ERNIE Bot (Baidu)",
      "name_zh": "文心一言 (百度)",
      "url": "https://yiyan.baidu.com",
      "country": "cn",
      "categories": [
        "chat",
        "image",
        "coding"
      ],
      "provider_page": "/en/providers/ernie/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "5 RPM",
      "api_notes": "Free credits via Baidu Cloud, ERNIE Speed/Lite are free",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "By Baidu, direct access in China, fast and enterprise-stable.",
      "models": [
        {
          "name": "ERNIE 4.5",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Baidu's latest flagship model"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "image",
        "api"
      ],
      "proof_url": "https://yiyan.baidu.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "fal",
      "name": "fal.ai",
      "name_zh": "fal.ai",
      "url": "https://fal.ai",
      "country": "us",
      "categories": [
        "image",
        "video"
      ],
      "provider_page": "/en/providers/fal/",
      "has_free_tier": true,
      "free_tier_limit": "Promotional credits on signup",
      "has_free_api": true,
      "api_free_credits": "Promotional credits",
      "api_rate_limit": "N/A",
      "api_notes": "Promotional credits on signup, pay-as-you-go after, image generation from $0.025/image",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Image generation is async, latency impact is minimal.",
      "models": [
        {
          "name": "Flux Dev",
          "context": "N/A",
          "rate_limit": "N/A",
          "notes": "High-quality image generation, $0.025/image"
        },
        {
          "name": "Stable Diffusion",
          "context": "N/A",
          "rate_limit": "N/A",
          "notes": "Open-source image generation"
        }
      ],
      "tags": [
        "image",
        "video",
        "api",
        "ai-generation"
      ],
      "proof_url": "https://fal.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "firered-openstoryline",
      "name": "FireRed-OpenStoryline",
      "name_zh": "FireRed-OpenStoryline",
      "url": "https://github.com/FireRedTeam/FireRed-OpenStoryline",
      "country": "cn",
      "categories": [
        "video",
        "agent",
        "editing"
      ],
      "provider_page": "/en/providers/firered-openstoryline/",
      "has_free_tier": true,
      "free_tier_limit": "Open-source self-hosting with no fixed usage limit",
      "has_free_api": true,
      "api_free_credits": "Unlimited self-hosted",
      "api_rate_limit": "Local/self-hosted",
      "api_notes": "Can run local Web/FastAPI/MCP services; model provider API keys must be configured in config.toml",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "GitHub/Hugging Face access may vary by network; Docker Hub and Alibaba Cloud image options are documented.",
      "models": [
        {
          "name": "OpenStoryline Video Editing Agent",
          "context": "N/A",
          "rate_limit": "Local/self-hosted",
          "notes": "Apache-2.0 open-source video editing agent with natural-language planning, MCP server, CLI, Web UI, and Docker deployment"
        }
      ],
      "tags": [
        "video",
        "open-source",
        "agent",
        "mcp",
        "self-hosted"
      ],
      "proof_url": "https://github.com/FireRedTeam/FireRed-OpenStoryline",
      "last_verified": "2026-05-10"
    },
    {
      "id": "fireworks",
      "name": "Fireworks AI",
      "name_zh": "Fireworks AI",
      "url": "https://fireworks.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/fireworks/",
      "has_free_tier": true,
      "free_tier_limit": "600 RPM",
      "has_free_api": true,
      "api_free_credits": "$1 free credits",
      "api_rate_limit": "600 RPM",
      "api_notes": "New users get $1 free credits, OpenAI-compatible API format",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Use openllmapi.com for direct China access.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "600 RPM",
          "notes": "Free to use, ultra-fast inference"
        },
        {
          "name": "Mixtral 8x22B",
          "context": "64k",
          "rate_limit": "600 RPM",
          "notes": "Large MoE model, available in free tier"
        }
      ],
      "tags": [
        "api",
        "fast-inference",
        "chat",
        "image"
      ],
      "proof_url": "https://fireworks.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "flux",
      "name": "FLUX (Black Forest Labs)",
      "name_zh": "FLUX (Black Forest Labs)",
      "url": "https://blackforestlabs.ai",
      "country": "de",
      "categories": [
        "image"
      ],
      "provider_page": "/en/providers/flux/",
      "has_free_tier": true,
      "free_tier_limit": "Free via third-party platforms",
      "has_free_api": true,
      "api_free_credits": "Free via platforms",
      "api_rate_limit": "Varies",
      "api_notes": "Free via Replicate/HuggingFace and other platforms",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "API platforms need proxy. Local deployment needs no network. ComfyUI is best local option.",
      "models": [
        {
          "name": "FLUX.1 [schnell]",
          "context": "N/A",
          "rate_limit": "Varies",
          "notes": "Open-source fast version, can run locally"
        },
        {
          "name": "FLUX.1 [dev]",
          "context": "N/A",
          "rate_limit": "Varies",
          "notes": "Open-source dev version, higher quality"
        }
      ],
      "tags": [
        "image",
        "open-source",
        "art",
        "self-hosted"
      ],
      "proof_url": "https://blackforestlabs.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "gemini",
      "name": "Gemini (Google)",
      "name_zh": "Gemini (Google)",
      "url": "https://gemini.google.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "image"
      ],
      "provider_page": "/en/providers/gemini/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "15 RPM",
      "api_notes": "Free API via Google AI Studio, 15 RPM",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "API accessible via proxy, AI Studio web requires proxy. Direct API access works in some regions but unstable.",
      "models": [
        {
          "name": "Gemini 2.5 Flash",
          "context": "1000k",
          "rate_limit": "Fair use",
          "notes": "Million-token context, fast"
        },
        {
          "name": "Gemini 2.5 Pro",
          "context": "1000k",
          "rate_limit": "Limited",
          "notes": "Limited for free users"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "coding",
        "long-context",
        "api"
      ],
      "proof_url": "https://gemini.google.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "google",
      "name": "Google AI (Gemini)",
      "name_zh": "Google AI (Gemini)",
      "url": "https://aistudio.google.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "multimodal"
      ],
      "provider_page": "/en/providers/google/",
      "has_free_tier": true,
      "free_tier_limit": "Gemini free tier unlimited",
      "has_free_api": true,
      "api_free_credits": "免费 API 无需信用卡",
      "api_rate_limit": "15 RPM (Flash)",
      "api_notes": "Google AI Studio offers completely free API access, no credit card required, rate limited",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Partially accessible, VPN recommended for stability",
      "models": [
        {
          "name": "Gemini 3.1 Pro",
          "context": "2M",
          "rate_limit": "2 RPM (free)",
          "notes": "Flagship model, 2M token context window"
        },
        {
          "name": "Gemini 3.1 Flash",
          "context": "1M",
          "rate_limit": "15 RPM (free)",
          "notes": "Fast model, 1M token context"
        },
        {
          "name": "Gemini 3.1 Flash Lite",
          "context": "1M",
          "rate_limit": "30 RPM (free)",
          "notes": "Lightest model, fastest speed"
        }
      ],
      "tags": [
        "llm",
        "api",
        "chat",
        "multimodal",
        "free",
        "coding"
      ],
      "proof_url": "https://aistudio.google.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "grok",
      "name": "Grok (xAI)",
      "name_zh": "Grok (xAI)",
      "url": "https://grok.x.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "image"
      ],
      "provider_page": "/en/providers/grok/",
      "has_free_tier": true,
      "free_tier_limit": "Limited requests/day",
      "has_free_api": true,
      "api_free_credits": "$25/月",
      "api_rate_limit": "Varies",
      "api_notes": "$25 free API credits on signup, resets monthly. Data sharing program ($150/month) ended May 2025.",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Both API and web require proxy. X account needed for registration.",
      "models": [
        {
          "name": "Grok 4.20",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "xAI's latest flagship, reasoning + non-reasoning, lowest hallucination rate"
        },
        {
          "name": "Grok-3",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "Free on X platform"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "api",
        "open-source",
        "reasoning"
      ],
      "proof_url": "https://grok.x.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "groq",
      "name": "Groq",
      "name_zh": "Groq",
      "url": "https://groq.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/groq/",
      "has_free_tier": true,
      "free_tier_limit": "6000 tokens/min (Llama 3.3 70B)",
      "has_free_api": true,
      "api_free_credits": "Free tier（永久免费）",
      "api_rate_limit": "30 RPM / 6000 TPM",
      "api_notes": "Free API powered by custom LPU (Language Processing Unit) chip, 10x+ faster than GPU.\nAPI keys start with gsk_. OpenAI-compatible format.\nFree tier has rate limits but no total cap, very generous for personal development.\n",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. API remains extremely fast even through proxy thanks to LPU chips. Use openllmapi.com as proxy.",
      "models": [
        {
          "name": "Llama 3.3 70B Versatile",
          "context": "128k",
          "rate_limit": "30 RPM / 6000 TPM",
          "notes": "World's fastest inference, 6000 tokens/min free, LPU chip accelerated"
        },
        {
          "name": "Llama 4 Scout 17B",
          "context": "128k",
          "rate_limit": "30 RPM / 6000 TPM",
          "notes": "Meta Llama 4 Scout, MoE architecture, free to use"
        },
        {
          "name": "Llama 4 Maverick 17B",
          "context": "128k",
          "rate_limit": "30 RPM / 6000 TPM",
          "notes": "Meta Llama 4 Maverick, MoE architecture, free to use"
        },
        {
          "name": "Mixtral 8x7B",
          "context": "32k",
          "rate_limit": "30 RPM / 5000 TPM",
          "notes": "MoE architecture, cost-effective"
        },
        {
          "name": "Gemma 2 9B",
          "context": "8k",
          "rate_limit": "30 RPM / 15000 TPM",
          "notes": "Google Gemma 2, ultra-fast small model"
        },
        {
          "name": "DeepSeek R1 Distill Llama 70B",
          "context": "128k",
          "rate_limit": "30 RPM / 6000 TPM",
          "notes": "DeepSeek R1 distilled, strong reasoning"
        }
      ],
      "tags": [
        "api",
        "fast-inference",
        "chat",
        "lpu",
        "free"
      ],
      "proof_url": "https://groq.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "huggingface",
      "name": "Hugging Face",
      "name_zh": "Hugging Face",
      "url": "https://huggingface.co",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/huggingface/",
      "has_free_tier": true,
      "free_tier_limit": "Varies by model",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "Varies",
      "api_notes": "Free Inference API tier, free Spaces hosting",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Accessible from China (proxy may be needed in some regions). hf-mirror.com is a Chinese mirror.",
      "models": [
        {
          "name": "Various Open Models",
          "context": "varies",
          "rate_limit": "Rate limited",
          "notes": "Thousands of free open-source models available"
        }
      ],
      "tags": [
        "platform",
        "open-source",
        "api",
        "community"
      ],
      "proof_url": "https://huggingface.co",
      "last_verified": "2026-05-10"
    },
    {
      "id": "spark",
      "name": "iFlytek Spark",
      "name_zh": "讯飞星火",
      "url": "https://xinghuo.xfyun.cn",
      "country": "cn",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/spark/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "200万 tokens",
      "api_rate_limit": "5 RPM",
      "api_notes": "2M tokens free on signup",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "By iFlytek, direct access in China, industry-leading voice capabilities.",
      "models": [
        {
          "name": "Spark 4.0 Ultra",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "iFlytek's strongest model, free to use"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "voice",
        "api"
      ],
      "proof_url": "https://xinghuo.xfyun.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "kimi",
      "name": "Kimi (Moonshot AI)",
      "name_zh": "Kimi (月之暗面)",
      "url": "https://kimi.moonshot.cn",
      "country": "cn",
      "categories": [
        "chat",
        "reasoning",
        "coding"
      ],
      "provider_page": "/en/providers/kimi/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "¥15 + 充 $5 送 $5",
      "api_rate_limit": "3 RPM",
      "api_notes": "¥15 free credits on signup. K2.5 API pricing $0.60/$2.50 per M tokens (input/output), top up $5 get $5 coupon.",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese service, direct access. Web version free and unlimited, one of the most popular AI assistants in China.",
      "models": [
        {
          "name": "Kimi-K2.5",
          "context": "1000k",
          "rate_limit": "Fair use",
          "notes": "Latest MoE model, 1M context, major coding/reasoning upgrade"
        },
        {
          "name": "Kimi-K2",
          "context": "200k",
          "rate_limit": "Fair use",
          "notes": "Ultra-long context, supports file parsing"
        }
      ],
      "tags": [
        "llm",
        "long-context",
        "chat",
        "coding",
        "moe"
      ],
      "proof_url": "https://kimi.moonshot.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "lmstudio",
      "name": "LM Studio",
      "name_zh": "LM Studio",
      "url": "https://lmstudio.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/lmstudio/",
      "has_free_tier": true,
      "free_tier_limit": "Unlimited (runs locally)",
      "has_free_api": true,
      "api_free_credits": "Unlimited",
      "api_rate_limit": "Local",
      "api_notes": "Local API server, OpenAI-compatible format",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Runs locally, no network needed. May need proxy to download software and models.",
      "models": [
        {
          "name": "多个开源模型",
          "context": "varies",
          "rate_limit": "Unlimited",
          "notes": "Runs locally, GUI interface, completely free"
        }
      ],
      "tags": [
        "local",
        "tool",
        "gui",
        "self-hosted"
      ],
      "proof_url": "https://lmstudio.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "minimax",
      "name": "MiniMax",
      "name_zh": "MiniMax (稀宇科技)",
      "url": "https://www.minimaxi.com",
      "country": "cn",
      "categories": [
        "chat",
        "audio",
        "video",
        "coding"
      ],
      "provider_page": "/en/providers/minimax/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "¥15",
      "api_rate_limit": "Varies",
      "api_notes": "¥15 free credits on signup. M2.7 is current SOTA for coding/tool-use. TTS API also has free tier.",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese platform, direct access, fast.",
      "models": [
        {
          "name": "MiniMax-M2.7",
          "context": "1000k",
          "rate_limit": "Fair use",
          "notes": "230B params, SOTA in coding/tool-use/search/office, 1M context"
        },
        {
          "name": "MiniMax-01",
          "context": "1000k",
          "rate_limit": "Fair use",
          "notes": "Million-token context, free to use"
        }
      ],
      "tags": [
        "llm",
        "tts",
        "audio",
        "long-context",
        "coding",
        "tool-use"
      ],
      "proof_url": "https://www.minimaxi.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "mistral",
      "name": "Mistral AI",
      "name_zh": "Mistral AI",
      "url": "https://chat.mistral.ai",
      "country": "fr",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/mistral/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "1 RPM",
      "api_notes": "Free API tier via La Plateforme",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. API works fine through proxy. Le Chat web also needs proxy.",
      "models": [
        {
          "name": "Mistral Large",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Europe's strongest model, free to use"
        },
        {
          "name": "Mistral Small",
          "context": "32k",
          "rate_limit": "Fair use",
          "notes": "Lightweight and fast, good for daily tasks"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "chat",
        "api",
        "european"
      ],
      "proof_url": "https://chat.mistral.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "novita",
      "name": "Novita AI",
      "name_zh": "Novita AI",
      "url": "https://novita.ai",
      "country": "us",
      "categories": [
        "chat",
        "image",
        "video"
      ],
      "provider_page": "/en/providers/novita/",
      "has_free_tier": true,
      "free_tier_limit": "credit-based",
      "has_free_api": true,
      "api_free_credits": "$0.50 free credits",
      "api_rate_limit": "60 RPM",
      "api_notes": "New users get $0.50 free credits, supports text and image generation",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Image generation API has demand in China, accessible via proxy.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "60 RPM",
          "notes": "Available in free tier"
        },
        {
          "name": "Stable Diffusion XL",
          "context": "N/A",
          "rate_limit": "60 RPM",
          "notes": "Image generation, available in free tier"
        }
      ],
      "tags": [
        "api",
        "chat",
        "image",
        "video"
      ],
      "proof_url": "https://novita.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "nvidia-build",
      "name": "NVIDIA Build (NIM API)",
      "name_zh": "NVIDIA Build (NIM API)",
      "url": "https://build.nvidia.com/",
      "country": "us",
      "categories": [
        "api",
        "chat",
        "coding",
        "reasoning",
        "multimodal"
      ],
      "provider_page": "/en/providers/nvidia-build/",
      "has_free_tier": true,
      "free_tier_limit": "Unlimited (40 RPM rate limit)",
      "has_free_api": true,
      "api_free_credits": "无限制（已取消额度限制）",
      "api_rate_limit": "40 RPM（可申请提升到 200 RPM）",
      "api_notes": "Free permanent API key on signup, 100+ models all free. Previous credit limits removed (was 1000 for personal, 5000 for enterprise email). OpenAI-compatible API, base_url is https://integrate.api.nvidia.com/v1. Direct access from China.",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Direct access from China via integrate.api.nvidia.com, no proxy needed. Medium speed, may slow during peak hours.",
      "models": [
        {
          "name": "MiniMax M2.7",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "230B params, coding/reasoning/office all-rounder"
        },
        {
          "name": "Kimi K2.5",
          "context": "1000k",
          "rate_limit": "40 RPM",
          "notes": "Moonshot native multimodal agentic model, 15T tokens training, 1M context, top Chinese ability"
        },
        {
          "name": "GLM-5.1",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "Zhipu's latest flagship, GLM-5 upgrade, optimized for agentic coding/long-horizon reasoning. GLM-5 deprecated 2026-04-20"
        },
        {
          "name": "DeepSeek V3.2",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "671B MoE, coding champion"
        },
        {
          "name": "DeepSeek R1",
          "context": "64k",
          "rate_limit": "40 RPM",
          "notes": "671B MoE, reasoning champion"
        },
        {
          "name": "Gemma 4 31B-IT",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "Google's latest open source, strong agentic capability, runs on consumer hardware"
        },
        {
          "name": "Nemotron-3-Super-120B",
          "context": "1000k",
          "rate_limit": "40 RPM",
          "notes": "NVIDIA's own flagship, hybrid Mamba-Transformer MoE, 1M context, 7.5x throughput vs Qwen3.5-122B"
        },
        {
          "name": "Llama 4 Maverick",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "Meta's latest open source LLM"
        },
        {
          "name": "Qwen 3.5",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "Alibaba Qwen, native multimodal, 397B params with only 17B active, extremely efficient"
        },
        {
          "name": "Step 3.5 Flash",
          "context": "128k",
          "rate_limit": "40 RPM",
          "notes": "StepFun, extremely fast"
        }
      ],
      "tags": [
        "api",
        "free",
        "multi-model",
        "llm",
        "nvidia",
        "unlimited"
      ],
      "proof_url": "https://build.nvidia.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "octoai",
      "name": "OctoAI",
      "name_zh": "OctoAI",
      "url": "https://octo.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/octoai/",
      "has_free_tier": true,
      "free_tier_limit": "credit-based",
      "has_free_api": true,
      "api_free_credits": "$10 free credits",
      "api_rate_limit": "60 RPM",
      "api_notes": "New users get $10 free credits, supports text and image generation",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Use openllmapi.com as proxy.",
      "models": [
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "60 RPM",
          "notes": "Available in free tier"
        },
        {
          "name": "SDXL",
          "context": "N/A",
          "rate_limit": "60 RPM",
          "notes": "Image generation, available in free tier"
        }
      ],
      "tags": [
        "api",
        "chat",
        "image"
      ],
      "proof_url": "https://octo.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "ollama",
      "name": "Ollama",
      "name_zh": "Ollama",
      "url": "https://ollama.com",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/ollama/",
      "has_free_tier": true,
      "free_tier_limit": "Unlimited (runs locally)",
      "has_free_api": true,
      "api_free_credits": "Unlimited",
      "api_rate_limit": "Local",
      "api_notes": "Local API, completely free, unlimited calls",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Runs locally, no network needed. May need proxy to download models (or use Chinese mirrors).",
      "models": [
        {
          "name": "Llama 3.3",
          "context": "128k",
          "rate_limit": "Unlimited",
          "notes": "Runs locally, completely free"
        },
        {
          "name": "Qwen2.5",
          "context": "32k",
          "rate_limit": "Unlimited",
          "notes": "Runs locally, completely free"
        },
        {
          "name": "DeepSeek-R1",
          "context": "64k",
          "rate_limit": "Unlimited",
          "notes": "Local reasoning model"
        }
      ],
      "tags": [
        "local",
        "open-source",
        "tool",
        "api",
        "self-hosted"
      ],
      "proof_url": "https://ollama.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "openai",
      "name": "OpenAI",
      "name_zh": "OpenAI",
      "url": "https://platform.openai.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image",
        "reasoning",
        "audio"
      ],
      "provider_page": "/en/providers/openai/",
      "has_free_tier": true,
      "free_tier_limit": "ChatGPT free tier unlimited",
      "has_free_api": true,
      "api_free_credits": "$5",
      "api_rate_limit": "3 RPM (free tier)",
      "api_notes": "$5 free credits for new accounts, valid for 3 months, pay-as-you-go after",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires VPN, not directly accessible from China",
      "models": [
        {
          "name": "GPT-5.2",
          "context": "128k",
          "rate_limit": "80 RPM",
          "notes": "Latest flagship model, strong multimodal capabilities"
        },
        {
          "name": "GPT-5.4 mini",
          "context": "128k",
          "rate_limit": "200 RPM",
          "notes": "Lightweight fast model, cost-effective"
        },
        {
          "name": "GPT Image 2",
          "context": "N/A",
          "rate_limit": "50 RPD",
          "notes": "Latest image generation model"
        }
      ],
      "tags": [
        "llm",
        "api",
        "chat",
        "image",
        "audio",
        "coding"
      ],
      "proof_url": "https://platform.openai.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "openrouter",
      "name": "OpenRouter",
      "name_zh": "OpenRouter",
      "url": "https://openrouter.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/openrouter/",
      "has_free_tier": true,
      "free_tier_limit": "Varies by model",
      "has_free_api": true,
      "api_free_credits": "Free models",
      "api_rate_limit": "20 RPM",
      "api_notes": "Some models are completely free, no payment needed",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Aggregates 400+ models, one key for all.",
      "models": [
        {
          "name": "Free Models (多个)",
          "context": "varies",
          "rate_limit": "20 RPM",
          "notes": "Multiple free models available (Llama, Gemma, etc.)"
        }
      ],
      "tags": [
        "api",
        "platform",
        "aggregator",
        "routing"
      ],
      "proof_url": "https://openrouter.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "qwen",
      "name": "Qwen (Alibaba)",
      "name_zh": "通义千问 (阿里)",
      "url": "https://tongyi.aliyun.com",
      "country": "cn",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "image",
        "multimodal"
      ],
      "provider_page": "/en/providers/qwen/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "7000 万 tokens（新用户一次性）",
      "api_rate_limit": "按模型不同",
      "api_notes": "Free credits via Alibaba Cloud Bailian. Qwen3.6-Plus priced at ¥2/M tokens (≤32K input).\nOpenRouter has free Preview (50/day default, 1000/day with $10+ credits).\nBailian Coding Plan Lite discontinued; qwen3.6-plus is Pro plan exclusive.\n",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Alibaba Cloud service, direct access in China, very fast. Top-tier Chinese understanding. Qwen3.6-Plus coding near Claude Sonnet level.",
      "models": [
        {
          "name": "Qwen3.6-Plus",
          "context": "1000k",
          "rate_limit": "Fair use",
          "notes": "Latest flagship, strongest Chinese coding model, near Claude Sonnet level"
        },
        {
          "name": "Qwen3.6-27B",
          "context": "262k (1M extended)",
          "rate_limit": "Open source",
          "notes": "27B dense model, open-source for local deployment, coding beats 397B models, supports Thinking Preservation"
        },
        {
          "name": "Qwen-Max",
          "context": "32k",
          "rate_limit": "Fair use",
          "notes": "Alibaba's strongest general model"
        },
        {
          "name": "Qwen-VL",
          "context": "32k",
          "rate_limit": "Fair use",
          "notes": "Multimodal, supports image understanding, OCR, object detection"
        }
      ],
      "tags": [
        "llm",
        "open-source",
        "chat",
        "coding",
        "multimodal",
        "reasoning"
      ],
      "proof_url": "https://tongyi.aliyun.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "replicate",
      "name": "Replicate",
      "name_zh": "Replicate",
      "url": "https://replicate.com",
      "country": "us",
      "categories": [
        "image",
        "video",
        "audio",
        "chat"
      ],
      "provider_page": "/en/providers/replicate/",
      "has_free_tier": true,
      "free_tier_limit": "Credit-based",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "Varies",
      "api_notes": "Monthly free inference credits",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Thousands of models, billed by compute time.",
      "models": [
        {
          "name": "FLUX.1",
          "context": "N/A",
          "rate_limit": "Rate limited",
          "notes": "Image generation model"
        },
        {
          "name": "Llama 3.3",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "Available within free credits"
        }
      ],
      "tags": [
        "platform",
        "image",
        "video",
        "audio",
        "api"
      ],
      "proof_url": "https://replicate.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "sambanova",
      "name": "SambaNova",
      "name_zh": "SambaNova",
      "url": "https://sambanova.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/sambanova/",
      "has_free_tier": true,
      "free_tier_limit": "30 RPM (no total cap)",
      "has_free_api": true,
      "api_free_credits": "Free tier（永久免费）",
      "api_rate_limit": "30 RPM",
      "api_notes": "Free API powered by custom RDU (Reconfigurable Dataflow Unit) chip, ultra-fast inference.\nAPI keys start with sn-. OpenAI-compatible format, base_url is https://api.sambanova.ai/v1.\nKey highlight: LLaMA 3.1 405B completely free — the only platform offering free 405B model access.\nFree tier has rate limits (30 RPM) but no total cap.\n",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Use openllmapi.com as proxy.",
      "models": [
        {
          "name": "Llama 3.1 405B",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "World's largest open-source model, completely free, RDU chip accelerated, ultra-fast"
        },
        {
          "name": "Llama 3.3 70B",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "Free to use, accelerated by custom RDU chip"
        },
        {
          "name": "DeepSeek R1",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "Reasoning model, free to use, 671B MoE"
        },
        {
          "name": "DeepSeek V3",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "DeepSeek V3 latest, free to use"
        },
        {
          "name": "Qwen 2.5 72B",
          "context": "128k",
          "rate_limit": "30 RPM",
          "notes": "Alibaba Qwen, free to use"
        }
      ],
      "tags": [
        "api",
        "fast-inference",
        "chat",
        "rdu",
        "free",
        "405b"
      ],
      "proof_url": "https://sambanova.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "sensenova",
      "name": "SenseNova",
      "name_zh": "商汤日日新 SenseNova",
      "url": "https://platform.sensenova.cn/?utm_source=yangmao.ai&utm_medium=referral&utm_campaign=sensenova_token_plan&utm_content=provider_profile",
      "country": "cn",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/sensenova/",
      "has_free_tier": true,
      "free_tier_limit": "Public beta Token Plan; quota requires login verification",
      "has_free_api": true,
      "api_free_credits": "公测免费额度",
      "api_rate_limit": "待官方控制台确认",
      "api_notes": "Phone signup reportedly unlocks a quick-start API key; endpoint reported as token.sensenova.cn/v1/chat/completions. Confirm in the official console.",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "China-based platform; expected to be China-friendly, but real availability, speed, and verification requirements need console validation.",
      "models": [
        {
          "name": "DeepSeek-V4-Flash",
          "context": "待官方控制台确认",
          "rate_limit": "待官方控制台确认",
          "notes": "A sourced article says SenseNova Token Plan provides free DeepSeek-V4-Flash access for document handling and low-cost API tests."
        }
      ],
      "tags": [
        "api",
        "china-access",
        "free-api",
        "deepseek",
        "sensetime"
      ],
      "proof_url": "https://platform.sensenova.cn/?utm_source=yangmao.ai&utm_medium=referral&utm_campaign=sensenova_token_plan&utm_content=provider_profile",
      "last_verified": "2026-05-10"
    },
    {
      "id": "siliconflow",
      "name": "SiliconFlow",
      "name_zh": "硅基流动 (SiliconFlow)",
      "url": "https://siliconflow.cn",
      "country": "cn",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/siliconflow/",
      "has_free_tier": true,
      "free_tier_limit": "Varies by model",
      "has_free_api": true,
      "api_free_credits": "¥14",
      "api_rate_limit": "Varies",
      "api_notes": "¥14 free credits on signup, some small models free forever",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Chinese service, direct access in China. Top choice for Chinese developers to access open-source model APIs.",
      "models": [
        {
          "name": "DeepSeek-V3 (hosted)",
          "context": "64k",
          "rate_limit": "Rate limited",
          "notes": "Free hosting of multiple open-source models"
        },
        {
          "name": "Qwen2.5-72B (hosted)",
          "context": "32k",
          "rate_limit": "Rate limited",
          "notes": "Free access to Qwen large model"
        }
      ],
      "tags": [
        "api",
        "platform",
        "inference"
      ],
      "proof_url": "https://siliconflow.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "stepfun",
      "name": "StepFun",
      "name_zh": "阶跃星辰",
      "url": "https://www.stepfun.com",
      "country": "cn",
      "categories": [
        "chat",
        "image",
        "video"
      ],
      "provider_page": "/en/providers/stepfun/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "¥10",
      "api_rate_limit": "5 RPM",
      "api_notes": "¥10 free credits on signup",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "StepFun, direct access in China. Step 3.5 Flash is extremely fast.",
      "models": [
        {
          "name": "Step-2",
          "context": "256k",
          "rate_limit": "Fair use",
          "notes": "Free to use via Yuewen"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "image",
        "video"
      ],
      "proof_url": "https://www.stepfun.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "hunyuan",
      "name": "Tencent Hunyuan",
      "name_zh": "腾讯混元",
      "url": "https://hunyuan.tencent.com",
      "country": "cn",
      "categories": [
        "chat",
        "image",
        "video"
      ],
      "provider_page": "/en/providers/hunyuan/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "100万 tokens",
      "api_rate_limit": "5 RPM",
      "api_notes": "Free credits via Tencent Cloud",
      "open_source": true,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "By Tencent, direct access in China, enterprise-stable.",
      "models": [
        {
          "name": "Hunyuan-Large",
          "context": "256k",
          "rate_limit": "Fair use",
          "notes": "Tencent's strongest model, free to use"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "image",
        "video",
        "open-source"
      ],
      "proof_url": "https://hunyuan.tencent.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "tiangong",
      "name": "Tiangong AI",
      "name_zh": "天工 AI (昆仑万维)",
      "url": "https://www.tiangong.cn",
      "country": "cn",
      "categories": [
        "chat",
        "search",
        "image"
      ],
      "provider_page": "/en/providers/tiangong/",
      "has_free_tier": true,
      "free_tier_limit": "No explicit limit",
      "has_free_api": true,
      "api_free_credits": "Free tier",
      "api_rate_limit": "Varies",
      "api_notes": "SkyWork API has free tier",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "By Kunlun Tech, direct access in China.",
      "models": [
        {
          "name": "天工大模型",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "AI search + chat, free to use"
        }
      ],
      "tags": [
        "llm",
        "search",
        "chat",
        "image"
      ],
      "proof_url": "https://www.tiangong.cn",
      "last_verified": "2026-05-10"
    },
    {
      "id": "together",
      "name": "Together AI",
      "name_zh": "Together AI",
      "url": "https://www.together.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image",
        "embedding"
      ],
      "provider_page": "/en/providers/together/",
      "has_free_tier": true,
      "free_tier_limit": "Credit-based ($5 signup bonus)",
      "has_free_api": true,
      "api_free_credits": "$5（注册赠送）",
      "api_rate_limit": "Varies by model",
      "api_notes": "$5 free credits on signup, supports 200+ open-source models.\nAPI keys start with together-. OpenAI-compatible, base_url is https://api.together.xyz/v1.\nKey highlight: FLUX.1 Schnell Free image generation is completely free, no credits consumed, unlimited!\nSupports LLM, image generation, embeddings, fine-tuning — full-stack AI capabilities.\n",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Moderate latency, rich model selection. Use openllmapi.com as proxy.",
      "models": [
        {
          "name": "Llama 3.3 70B Turbo",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "Available within free credits, Turbo accelerated"
        },
        {
          "name": "Llama 4 Maverick",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "Meta's latest Llama 4, MoE architecture"
        },
        {
          "name": "Mixtral 8x22B",
          "context": "64k",
          "rate_limit": "Rate limited",
          "notes": "Available within free credits"
        },
        {
          "name": "DeepSeek V3",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "DeepSeek V3, available within free credits"
        },
        {
          "name": "FLUX.1 Schnell (Free)",
          "context": "",
          "rate_limit": "无限制",
          "notes": "FLUX image generation completely free, no credits consumed! Unlimited use"
        },
        {
          "name": "FLUX.1 Schnell",
          "context": "",
          "rate_limit": "Rate limited",
          "notes": "FLUX high-quality version, consumes credits"
        }
      ],
      "tags": [
        "api",
        "platform",
        "inference",
        "open-source",
        "image",
        "flux",
        "free"
      ],
      "proof_url": "https://www.together.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "vidu",
      "name": "Vidu",
      "name_zh": "Vidu",
      "url": "https://www.vidu.com",
      "country": "cn",
      "categories": [
        "video",
        "image"
      ],
      "provider_page": "/en/providers/vidu/",
      "has_free_tier": true,
      "free_tier_limit": "80 credits/month",
      "has_free_api": true,
      "api_free_credits": "$1",
      "api_rate_limit": "N/A",
      "api_notes": "$1 free credit on API signup (~14 short clips), unlimited free in Off-Peak mode",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Chinese platform, direct access in China, fast speed.",
      "models": [
        {
          "name": "Vidu Q3",
          "context": "N/A",
          "rate_limit": "N/A",
          "notes": "AI video generation, ~5 credits per 4s clip"
        }
      ],
      "tags": [
        "video",
        "image",
        "ai-generation"
      ],
      "proof_url": "https://www.vidu.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "aider",
      "name": "Aider",
      "name_zh": "Aider",
      "url": "https://aider.chat/",
      "country": "",
      "categories": [
        "coding",
        "cli"
      ],
      "provider_page": "/en/providers/aider/",
      "has_free_tier": true,
      "free_tier_limit": "MIT open-source; bring your own model API key, pay-per-use.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://aider.chat/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "brave-search-api",
      "name": "Brave Search API",
      "name_zh": "Brave Search API",
      "url": "https://brave.com/search/api/",
      "country": "",
      "categories": [
        "search",
        "api"
      ],
      "provider_page": "/en/providers/brave-search-api/",
      "has_free_tier": true,
      "free_tier_limit": "2000 free queries per month.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://brave.com/search/api/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cerebras-cloud",
      "name": "Cerebras Cloud",
      "name_zh": "Cerebras Cloud",
      "url": "https://cloud.cerebras.ai/",
      "country": "",
      "categories": [
        "api",
        "inference"
      ],
      "provider_page": "/en/providers/cerebras-cloud/",
      "has_free_tier": true,
      "free_tier_limit": "Free developer tier at 30 RPM.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://cloud.cerebras.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "chatgpt",
      "name": "ChatGPT (OpenAI)",
      "name_zh": "ChatGPT (OpenAI)",
      "url": "https://chat.openai.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "image",
        "agent"
      ],
      "provider_page": "/en/providers/chatgpt/",
      "has_free_tier": true,
      "free_tier_limit": "Limited requests/day",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment, no free tier. Codex (cloud coding agent) now available for enterprises.",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Recommend API relay services or self-hosted Cloudflare Workers proxy.",
      "models": [
        {
          "name": "GPT-5.5",
          "context": "1000k",
          "rate_limit": "Plus/Pro only",
          "notes": "Released April 2026, smartest model, major upgrade for coding/research/analysis. API $5/M input, $30/M output"
        },
        {
          "name": "GPT-5.5-Cyber",
          "context": "1000k",
          "rate_limit": "Trusted Access / enterprise-limited",
          "notes": "GPT-5.5 variant for trusted-access cybersecurity use cases, including security evaluation, vulnerability analysis, and defensive workflows; availability is gated through OpenAI trusted access."
        },
        {
          "name": "GPT-5.5 Pro",
          "context": "1050k",
          "rate_limit": "Pro only",
          "notes": "Deep reasoning variant for complex high-stakes tasks. API $30/M input, $180/M output"
        },
        {
          "name": "GPT-5.4",
          "context": "128k",
          "rate_limit": "Plus/Pro",
          "notes": "Previous flagship, still powerful"
        },
        {
          "name": "GPT-4o mini",
          "context": "128k",
          "rate_limit": "Rate limited",
          "notes": "Available to free users, fast"
        },
        {
          "name": "GPT-4o",
          "context": "128k",
          "rate_limit": "Very limited",
          "notes": "Very limited for free users"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "image",
        "coding",
        "agent"
      ],
      "proof_url": "https://chat.openai.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "claude",
      "name": "Claude (Anthropic)",
      "name_zh": "Claude (Anthropic)",
      "url": "https://claude.ai",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning"
      ],
      "provider_page": "/en/providers/claude/",
      "has_free_tier": true,
      "free_tier_limit": "Limited messages/day",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment, no free tier",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. API accessible via relay services. Registration needs non-China phone number.",
      "models": [
        {
          "name": "Claude Sonnet 4.5",
          "context": "200k",
          "rate_limit": "Rate limited",
          "notes": "Available to free users, excellent at coding"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "coding",
        "long-context"
      ],
      "proof_url": "https://claude.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cline",
      "name": "Cline",
      "name_zh": "Cline",
      "url": "https://cline.bot/",
      "country": "",
      "categories": [
        "coding",
        "agent"
      ],
      "provider_page": "/en/providers/cline/",
      "has_free_tier": true,
      "free_tier_limit": "Free and open-source extension; plug in DeepSeek/Qwen for near-zero cost.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://cline.bot/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "codeium",
      "name": "Codeium",
      "name_zh": "Codeium",
      "url": "https://codeium.com/",
      "country": "",
      "categories": [
        "coding",
        "autocomplete"
      ],
      "provider_page": "/en/providers/codeium/",
      "has_free_tier": true,
      "free_tier_limit": "Free forever for individuals with unlimited autocomplete and chat.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://codeium.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "continue",
      "name": "Continue",
      "name_zh": "Continue",
      "url": "https://continue.dev/",
      "country": "",
      "categories": [
        "coding",
        "agent"
      ],
      "provider_page": "/en/providers/continue/",
      "has_free_tier": true,
      "free_tier_limit": "Apache-2.0 open-source, free. Pairs with local Ollama for zero-cost offline use.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://continue.dev/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "cursor",
      "name": "Cursor",
      "name_zh": "Cursor",
      "url": "https://cursor.com",
      "country": "us",
      "categories": [
        "coding"
      ],
      "provider_page": "/en/providers/cursor/",
      "has_free_tier": true,
      "free_tier_limit": "2000 completions + 50 premium requests/mo",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No standalone API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Cursor IDE accessible from China. AI features routed through Cursor servers.",
      "models": [
        {
          "name": "Cursor (多模型)",
          "context": "varies",
          "rate_limit": "50 premium/month",
          "notes": "50 premium requests per month on free tier"
        }
      ],
      "tags": [
        "coding",
        "ide",
        "editor"
      ],
      "proof_url": "https://cursor.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "d-id",
      "name": "D-ID",
      "name_zh": "D-ID",
      "url": "https://www.d-id.com/",
      "country": "",
      "categories": [
        "video",
        "avatar"
      ],
      "provider_page": "/en/providers/d-id/",
      "has_free_tier": true,
      "free_tier_limit": "5 free minutes on signup.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.d-id.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "databricks-mosaic",
      "name": "Databricks Mosaic AI",
      "name_zh": "Databricks Mosaic AI",
      "url": "https://www.databricks.com/product/machine-learning/mosaic-ai",
      "country": "",
      "categories": [
        "api",
        "enterprise"
      ],
      "provider_page": "/en/providers/databricks-mosaic/",
      "has_free_tier": true,
      "free_tier_limit": "$400 free credits for new accounts.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.databricks.com/product/machine-learning/mosaic-ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "github-copilot",
      "name": "GitHub Copilot",
      "name_zh": "GitHub Copilot",
      "url": "https://github.com/features/copilot",
      "country": "us",
      "categories": [
        "coding"
      ],
      "provider_page": "/en/providers/github-copilot/",
      "has_free_tier": true,
      "free_tier_limit": "2000 completions/mo + 50 chat messages/mo",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No standalone API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Direct access from China. Supports VS Code and JetBrains.",
      "models": [
        {
          "name": "Copilot (GPT-4o based)",
          "context": "8k",
          "rate_limit": "2000/month",
          "notes": "Free tier: 2000 code completions per month"
        }
      ],
      "tags": [
        "coding",
        "ide"
      ],
      "proof_url": "https://github.com/features/copilot",
      "last_verified": "2026-05-10"
    },
    {
      "id": "heygen",
      "name": "HeyGen",
      "name_zh": "HeyGen",
      "url": "https://www.heygen.com/",
      "country": "",
      "categories": [
        "video",
        "avatar"
      ],
      "provider_page": "/en/providers/heygen/",
      "has_free_tier": true,
      "free_tier_limit": "3 free generations/mo; Creator $29/mo.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.heygen.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "hfviewer",
      "name": "HFViewer",
      "name_zh": "HFViewer",
      "url": "https://hfviewer.com",
      "country": "se",
      "categories": [
        "chat"
      ],
      "provider_page": "/en/providers/hfviewer/",
      "has_free_tier": true,
      "free_tier_limit": "Free web access",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No public API or API pricing found; the core product is a free web visualizer",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "The site is directly accessible; model metadata depends on Hugging Face and may require a proxy in some regions.",
      "models": [
        {
          "name": "Hugging Face Model Visualizer",
          "context": "N/A",
          "rate_limit": "未公开",
          "notes": "Paste a Hugging Face model URL or repo name to inspect an interactive architecture graph"
        }
      ],
      "tags": [
        "tool",
        "model-visualizer",
        "huggingface",
        "model-analysis"
      ],
      "proof_url": "https://hfviewer.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "ideogram",
      "name": "Ideogram",
      "name_zh": "Ideogram",
      "url": "https://ideogram.ai",
      "country": "us",
      "categories": [
        "image"
      ],
      "provider_page": "/en/providers/ideogram/",
      "has_free_tier": true,
      "free_tier_limit": "10 images/day",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Best text rendering in the industry.",
      "models": [
        {
          "name": "Ideogram 3.0",
          "context": "N/A",
          "rate_limit": "10/day",
          "notes": "Best text rendering in image generation"
        }
      ],
      "tags": [
        "image",
        "art",
        "text-rendering"
      ],
      "proof_url": "https://ideogram.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "jan",
      "name": "Jan",
      "name_zh": "Jan",
      "url": "https://jan.ai/",
      "country": "",
      "categories": [
        "selfhosted",
        "desktop"
      ],
      "provider_page": "/en/providers/jan/",
      "has_free_tier": true,
      "free_tier_limit": "AGPL-3.0 open-source, free.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://jan.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "jina-ai",
      "name": "Jina AI",
      "name_zh": "Jina AI",
      "url": "https://jina.ai/",
      "country": "",
      "categories": [
        "embedding",
        "api"
      ],
      "provider_page": "/en/providers/jina-ai/",
      "has_free_tier": true,
      "free_tier_limit": "1M free tokens on signup.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://jina.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "kling",
      "name": "Kling AI (Kuaishou)",
      "name_zh": "可灵 AI (快手)",
      "url": "https://klingai.com",
      "country": "cn",
      "categories": [
        "video",
        "image"
      ],
      "provider_page": "/en/providers/kling/",
      "has_free_tier": true,
      "free_tier_limit": "66 credits/day",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "By Kuaishou, direct access in China, fast. Leading video generation quality domestically.",
      "models": [
        {
          "name": "Kling 2.0",
          "context": "N/A",
          "rate_limit": "66 credits/day",
          "notes": "AI video generation, 66 credits/day free"
        }
      ],
      "tags": [
        "video",
        "image",
        "creative"
      ],
      "proof_url": "https://klingai.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "localai",
      "name": "LocalAI",
      "name_zh": "LocalAI",
      "url": "https://localai.io/",
      "country": "",
      "categories": [
        "selfhosted",
        "api"
      ],
      "provider_page": "/en/providers/localai/",
      "has_free_tier": true,
      "free_tier_limit": "MIT open-source, zero API cost when self-hosted.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://localai.io/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "luma",
      "name": "Luma (Dream Machine)",
      "name_zh": "Luma (Dream Machine)",
      "url": "https://lumalabs.ai",
      "country": "us",
      "categories": [
        "video",
        "3d"
      ],
      "provider_page": "/en/providers/luma/",
      "has_free_tier": true,
      "free_tier_limit": "Free trial credits",
      "has_free_api": false,
      "api_free_credits": "免费试用",
      "api_rate_limit": "N/A",
      "api_notes": "Free credits on web, API requires separate credit purchase",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy in China. Video generation is async, proxy impact is minimal.",
      "models": [
        {
          "name": "Dream Machine",
          "context": "N/A",
          "rate_limit": "N/A",
          "notes": "AI video generation, supports 4K and audio"
        }
      ],
      "tags": [
        "video",
        "3d",
        "ai-generation"
      ],
      "proof_url": "https://lumalabs.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "meshy",
      "name": "Meshy",
      "name_zh": "Meshy",
      "url": "https://www.meshy.ai/",
      "country": "",
      "categories": [
        "3d",
        "generation"
      ],
      "provider_page": "/en/providers/meshy/",
      "has_free_tier": true,
      "free_tier_limit": "200 free credits per month.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.meshy.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "midjourney",
      "name": "Midjourney",
      "name_zh": "Midjourney",
      "url": "https://www.midjourney.com",
      "country": "us",
      "categories": [
        "image"
      ],
      "provider_page": "/en/providers/midjourney/",
      "has_free_tier": true,
      "free_tier_limit": "25 images/day (limited time)",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No free API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy and Discord account. High barrier for China users.",
      "models": [
        {
          "name": "Midjourney v7",
          "context": "N/A",
          "rate_limit": "25/day",
          "notes": "Free trial on web"
        }
      ],
      "tags": [
        "image",
        "art",
        "creative"
      ],
      "proof_url": "https://www.midjourney.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "millionengine",
      "name": "Million Engine",
      "name_zh": "万量引擎",
      "url": "https://millionengine.com",
      "country": "cn",
      "categories": [
        "api",
        "platform"
      ],
      "provider_page": "/en/providers/millionengine/",
      "has_free_tier": false,
      "free_tier_limit": "No free tier",
      "has_free_api": false,
      "api_free_credits": "按量付费",
      "api_rate_limit": "N/A",
      "api_notes": "Pay-as-you-go, near-official pricing, some models discounted",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": true,
      "china_access_notes": "Direct China access, optimized for Chinese developers, low latency",
      "models": [],
      "tags": [
        "api",
        "aggregator",
        "platform",
        "china",
        "multimodal"
      ],
      "proof_url": "https://millionengine.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "notion-ai",
      "name": "Notion AI",
      "name_zh": "Notion AI",
      "url": "https://www.notion.so",
      "country": "us",
      "categories": [
        "chat",
        "coding"
      ],
      "provider_page": "/en/providers/notion-ai/",
      "has_free_tier": true,
      "free_tier_limit": "Limited requests",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No standalone API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Notion accessible from China (occasionally unstable in some regions).",
      "models": [
        {
          "name": "Notion AI",
          "context": "varies",
          "rate_limit": "Limited",
          "notes": "AI assistant integrated in Notion"
        }
      ],
      "tags": [
        "writing",
        "productivity",
        "chat"
      ],
      "proof_url": "https://www.notion.so",
      "last_verified": "2026-05-10"
    },
    {
      "id": "nvidia-nim",
      "name": "NVIDIA NIM",
      "name_zh": "NVIDIA NIM",
      "url": "https://build.nvidia.com/",
      "country": "",
      "categories": [
        "api",
        "inference"
      ],
      "provider_page": "/en/providers/nvidia-nim/",
      "has_free_tier": true,
      "free_tier_limit": "1000 free requests for developers.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://build.nvidia.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "onepass",
      "name": "OnePass (Google AI Pro)",
      "name_zh": "OnePass (Google AI Pro)",
      "url": "https://onepass.fun",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "reasoning",
        "multimodal"
      ],
      "provider_page": "/en/providers/onepass/",
      "has_free_tier": true,
      "free_tier_limit": "Google AI Pro free for one year",
      "has_free_api": false,
      "api_free_credits": "Google AI Pro 免费一年（绑 Visa 卡不扣费）",
      "api_rate_limit": "Google AI Pro 标准限制",
      "api_notes": "Register Google AI Pro via OnePass platform, bind Visa card without actual charges, free for one year.\nGet full Google AI Pro features including Gemini Pro series models.\nNote: This is a Google AI Pro subscription, not a standalone API platform.\nRequires a Visa credit/debit card for verification, but no actual charges.\n",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Google services require proxy in China. OnePass registration may also need proxy.",
      "models": [
        {
          "name": "Gemini Pro",
          "context": "1000k",
          "rate_limit": "Google AI Pro 标准限制",
          "notes": "Full Google Gemini Pro access, free for one year via OnePass"
        },
        {
          "name": "Gemini Pro Vision",
          "context": "1000k",
          "rate_limit": "Google AI Pro 标准限制",
          "notes": "Gemini Pro multimodal, supports image understanding"
        }
      ],
      "tags": [
        "google",
        "gemini",
        "free",
        "subscription",
        "multimodal"
      ],
      "proof_url": "https://onepass.fun",
      "last_verified": "2026-05-10"
    },
    {
      "id": "sora",
      "name": "OpenAI Sora",
      "name_zh": "OpenAI Sora",
      "url": "https://sora.com/",
      "country": "us",
      "categories": [
        "video",
        "image"
      ],
      "provider_page": "/en/providers/sora/",
      "has_free_tier": false,
      "free_tier_limit": "No standalone free tier; usually bundled with ChatGPT Plus/Pro benefits",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "Sora is primarily a product surface; public API/free API availability depends on OpenAI's latest docs.",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "OpenAI services usually require proxy and non-mainland payment/account setup in mainland China.",
      "models": [
        {
          "name": "Sora",
          "context": "N/A",
          "rate_limit": "Plan-based",
          "notes": "OpenAI video generation model with plan-based quota, watermark and resolution limits"
        }
      ],
      "tags": [
        "video",
        "openai",
        "text-to-video",
        "creative"
      ],
      "proof_url": "https://sora.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "perplexity",
      "name": "Perplexity AI",
      "name_zh": "Perplexity AI",
      "url": "https://www.perplexity.ai",
      "country": "us",
      "categories": [
        "search",
        "chat",
        "reasoning"
      ],
      "provider_page": "/en/providers/perplexity/",
      "has_free_tier": true,
      "free_tier_limit": "No limit (basic search)",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment",
      "open_source": false,
      "openai_compatible": true,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Perplexity's strength is web search and real-time information.",
      "models": [
        {
          "name": "Perplexity Default",
          "context": "32k",
          "rate_limit": "Fair use",
          "notes": "AI search engine, real-time web access"
        },
        {
          "name": "Pro Search",
          "context": "32k",
          "rate_limit": "5/day free",
          "notes": "Deep search, 5 per day for free users"
        }
      ],
      "tags": [
        "search",
        "chat",
        "web"
      ],
      "proof_url": "https://www.perplexity.ai",
      "last_verified": "2026-05-10"
    },
    {
      "id": "pika",
      "name": "Pika Labs",
      "name_zh": "Pika Labs",
      "url": "https://pika.art/",
      "country": "",
      "categories": [
        "video",
        "generation"
      ],
      "provider_page": "/en/providers/pika/",
      "has_free_tier": true,
      "free_tier_limit": "Daily free credits; Standard $10/mo.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://pika.art/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "playht",
      "name": "PlayHT",
      "name_zh": "PlayHT",
      "url": "https://play.ht/",
      "country": "",
      "categories": [
        "voice",
        "tts"
      ],
      "provider_page": "/en/providers/playht/",
      "has_free_tier": true,
      "free_tier_limit": "12.5K chars/mo free tier.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://play.ht/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "poe",
      "name": "Poe",
      "name_zh": "Poe",
      "url": "https://poe.com",
      "country": "us",
      "categories": [
        "chat",
        "coding",
        "image"
      ],
      "provider_page": "/en/providers/poe/",
      "has_free_tier": true,
      "free_tier_limit": "Credit-based, daily refresh",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No free API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. By Quora, aggregates multiple AI models.",
      "models": [
        {
          "name": "GPT-4o mini",
          "context": "128k",
          "rate_limit": "Unlimited",
          "notes": "Free unlimited access"
        },
        {
          "name": "Claude Sonnet",
          "context": "200k",
          "rate_limit": "Limited credits",
          "notes": "Uses credits, limited daily"
        },
        {
          "name": "Gemini Pro",
          "context": "128k",
          "rate_limit": "Limited credits",
          "notes": "Uses credits"
        }
      ],
      "tags": [
        "platform",
        "aggregator",
        "chat"
      ],
      "proof_url": "https://poe.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "red-note-ai",
      "name": "Red Note AI",
      "name_zh": "红笔记AI",
      "url": "https://www.red-note.top/zh",
      "country": "cn",
      "categories": [
        "writing",
        "marketing",
        "social"
      ],
      "provider_page": "/en/providers/red-note-ai/",
      "has_free_tier": true,
      "free_tier_limit": "First 3 generations free",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No public API or developer pricing found; free quota applies to the web app's first 3 generations",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "The website is directly accessible and focused on Chinese content creators.",
      "models": [
        {
          "name": "多AI小红书笔记生成",
          "context": "N/A",
          "rate_limit": "3 free generations initially",
          "notes": "Persona-based Xiaohongshu content generation with viral-post references, content management, and multiple AI engines"
        },
        {
          "name": "通义千问 / 豆包 / DeepSeek / GPT-5.1 / Gemini 3.0",
          "context": "varies",
          "rate_limit": "按套餐不同",
          "notes": "Official FAQ mentions multiple Chinese and global AI engines, recommended by content type"
        }
      ],
      "tags": [
        "writing",
        "xiaohongshu",
        "rednote",
        "persona",
        "free-trial"
      ],
      "proof_url": "https://www.red-note.top/zh",
      "last_verified": "2026-05-10"
    },
    {
      "id": "runway",
      "name": "Runway",
      "name_zh": "Runway",
      "url": "https://runwayml.com",
      "country": "us",
      "categories": [
        "video",
        "image"
      ],
      "provider_page": "/en/providers/runway/",
      "has_free_tier": true,
      "free_tier_limit": "125 credits (one-time)",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "API requires payment",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Industry-leading AI video generation quality.",
      "models": [
        {
          "name": "Gen-3 Alpha",
          "context": "N/A",
          "rate_limit": "Credit-based",
          "notes": "AI video generation, 125 credits on signup"
        }
      ],
      "tags": [
        "video",
        "image",
        "creative"
      ],
      "proof_url": "https://runwayml.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "seedance",
      "name": "Seedance (Doubao Video)",
      "name_zh": "Seedance（豆包视频）",
      "url": "https://www.doubao.com/chat/create-video",
      "country": "cn",
      "categories": [
        "video",
        "image"
      ],
      "provider_page": "/en/providers/seedance/",
      "has_free_tier": true,
      "free_tier_limit": "Free generation credits vary by Doubao campaign and account",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "Listed primarily as Doubao's consumer video creation entry; API access depends on Volcengine/Doubao enterprise offerings.",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Doubao's China entry is directly accessible and suitable for Chinese creators.",
      "models": [
        {
          "name": "Seedance 2.0",
          "context": "N/A",
          "rate_limit": "Credit-based",
          "notes": "ByteDance video generation model for text-to-video, image-to-video and short-form creative assets"
        }
      ],
      "tags": [
        "video",
        "doubao",
        "bytedance",
        "text-to-video"
      ],
      "proof_url": "https://www.doubao.com/chat/create-video",
      "last_verified": "2026-05-10"
    },
    {
      "id": "stability-ai",
      "name": "Stability AI",
      "name_zh": "Stability AI",
      "url": "https://platform.stability.ai/",
      "country": "",
      "categories": [
        "image",
        "api"
      ],
      "provider_page": "/en/providers/stability-ai/",
      "has_free_tier": true,
      "free_tier_limit": "25 free credits on signup.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://platform.stability.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "suno",
      "name": "Suno",
      "name_zh": "Suno",
      "url": "https://suno.com",
      "country": "us",
      "categories": [
        "audio"
      ],
      "provider_page": "/en/providers/suno/",
      "has_free_tier": true,
      "free_tier_limit": "50 credits/day (~10 songs)",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No free API",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "Requires proxy. Best AI music generation quality.",
      "models": [
        {
          "name": "Suno v4",
          "context": "N/A",
          "rate_limit": "50 credits/day",
          "notes": "AI music generation, ~10 songs/day free"
        }
      ],
      "tags": [
        "audio",
        "music",
        "creative"
      ],
      "proof_url": "https://suno.com",
      "last_verified": "2026-05-10"
    },
    {
      "id": "supermaven",
      "name": "Supermaven",
      "name_zh": "Supermaven",
      "url": "https://supermaven.com/",
      "country": "",
      "categories": [
        "coding",
        "autocomplete"
      ],
      "provider_page": "/en/providers/supermaven/",
      "has_free_tier": true,
      "free_tier_limit": "Free basic tier; Pro $10/mo.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://supermaven.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "synthesia",
      "name": "Synthesia",
      "name_zh": "Synthesia",
      "url": "https://www.synthesia.io/",
      "country": "",
      "categories": [
        "video",
        "avatar"
      ],
      "provider_page": "/en/providers/synthesia/",
      "has_free_tier": true,
      "free_tier_limit": "3 min/mo free; Starter $22/mo for 120 min.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.synthesia.io/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "tabby",
      "name": "Tabby",
      "name_zh": "Tabby",
      "url": "https://tabby.tabbyml.com/",
      "country": "",
      "categories": [
        "coding",
        "selfhosted"
      ],
      "provider_page": "/en/providers/tabby/",
      "has_free_tier": true,
      "free_tier_limit": "Apache-2.0 open-source, self-host for zero API cost.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://tabby.tabbyml.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "yuanbao",
      "name": "Tencent Yuanbao",
      "name_zh": "腾讯元宝",
      "url": "https://yuanbao.tencent.com/",
      "country": "cn",
      "categories": [
        "chat",
        "search",
        "reasoning"
      ],
      "provider_page": "/en/providers/yuanbao/",
      "has_free_tier": true,
      "free_tier_limit": "Free web/app access; exact quota changes dynamically",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "Yuanbao is a consumer AI assistant. API access usually goes through Tencent Cloud Hunyuan/DeepSeek APIs, not Yuanbao itself.",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "Tencent-hosted China service with fast direct access in mainland China.",
      "models": [
        {
          "name": "Tencent Hunyuan",
          "context": "128k",
          "rate_limit": "Fair use",
          "notes": "Tencent Hunyuan entry point for Chinese Q&A, search summaries and office writing"
        },
        {
          "name": "DeepSeek R1/V3 access",
          "context": "Varies",
          "rate_limit": "Fair use",
          "notes": "Yuanbao exposes DeepSeek-related capabilities inside the app, useful for China LLM comparisons"
        }
      ],
      "tags": [
        "llm",
        "chat",
        "chinese-ai",
        "search"
      ],
      "proof_url": "https://yuanbao.tencent.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "tripo",
      "name": "Tripo 3D",
      "name_zh": "Tripo 3D",
      "url": "https://www.tripo3d.ai/",
      "country": "",
      "categories": [
        "3d",
        "generation"
      ],
      "provider_page": "/en/providers/tripo/",
      "has_free_tier": true,
      "free_tier_limit": "100 free credits daily.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.tripo3d.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "udio",
      "name": "Udio",
      "name_zh": "Udio",
      "url": "https://www.udio.com/",
      "country": "",
      "categories": [
        "music",
        "generation"
      ],
      "provider_page": "/en/providers/udio/",
      "has_free_tier": true,
      "free_tier_limit": "1200 generations/mo free; Pro $10/mo.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.udio.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "vllm",
      "name": "vLLM",
      "name_zh": "vLLM",
      "url": "https://docs.vllm.ai/",
      "country": "",
      "categories": [
        "selfhosted",
        "inference"
      ],
      "provider_page": "/en/providers/vllm/",
      "has_free_tier": true,
      "free_tier_limit": "Apache-2.0 open-source.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": true,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://docs.vllm.ai/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "voyage-ai",
      "name": "Voyage AI",
      "name_zh": "Voyage AI",
      "url": "https://www.voyageai.com/",
      "country": "",
      "categories": [
        "embedding",
        "api"
      ],
      "provider_page": "/en/providers/voyage-ai/",
      "has_free_tier": true,
      "free_tier_limit": "50M free tokens for new users.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://www.voyageai.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "windsurf",
      "name": "Windsurf",
      "name_zh": "Windsurf",
      "url": "https://windsurf.com/",
      "country": "",
      "categories": [
        "coding",
        "ide"
      ],
      "provider_page": "/en/providers/windsurf/",
      "has_free_tier": true,
      "free_tier_limit": "Free plan with Cascade Base credits; Pro $15/mo unlocks premium models.",
      "has_free_api": false,
      "api_free_credits": "",
      "api_rate_limit": "",
      "api_notes": "",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": false,
      "china_access_notes": "",
      "models": [],
      "tags": [],
      "proof_url": "https://windsurf.com/",
      "last_verified": "2026-05-10"
    },
    {
      "id": "xiaobiji",
      "name": "Xiaobiji",
      "name_zh": "小笔记",
      "url": "https://xiaobiji.app",
      "country": "cn",
      "categories": [
        "writing",
        "marketing",
        "social"
      ],
      "provider_page": "/en/providers/xiaobiji/",
      "has_free_tier": true,
      "free_tier_limit": "Free web tool; exact usage limit not disclosed",
      "has_free_api": false,
      "api_free_credits": "$0",
      "api_rate_limit": "N/A",
      "api_notes": "No public API or API pricing found; official metadata describes it as a free online writing tool",
      "open_source": false,
      "openai_compatible": false,
      "china_direct": true,
      "china_access_notes": "The website is directly accessible; generation availability depends on the live web app.",
      "models": [
        {
          "name": "Xiaohongshu Post Generator",
          "context": "N/A",
          "rate_limit": "未公开",
          "notes": "AI post generator for Xiaohongshu/RedNote creators with topic ideas, 8 content templates, and sensitive-word checking"
        }
      ],
      "tags": [
        "writing",
        "xiaohongshu",
        "rednote",
        "social-media",
        "free-tool"
      ],
      "proof_url": "https://xiaobiji.app",
      "last_verified": "2026-05-10"
    }
  ]
}