431+ 免费大模型 API，来自 30 个提供商 — 一站式发现、对比、配置免费模型。
🌐 访问 freellm.net — 浏览模型 · 在线体验 · 配置生成器 · API 密钥
🔄 数据每日从 freellm.net 自动更新 — 最后更新: 2026-08-12
🌐 English · 简体中文 · 繁體中文 · 日本語 · 한국어
--- ## 为什么需要这个项目 找一个免费 LLM API，不应该翻十几个 GitHub README、注册五个不同平台、或者猜测哪些模型还有免费额度。 这个仓库是一个**结构化、机器可读的免费 LLM API 目录** — 包含速率限制、上下文窗口、一键配置代码片段和直达 API 密钥链接。每日更新。 **为什么选择这个仓库 + freellm.net：** - ✅ **始终最新** — 数据每日自动化监控更新，不是两年前的静态列表 - ✅ **信用卡透明** — 清晰标注哪些提供商需要信用卡、手机验证，哪些完全无需 - ✅ **一键配置** — 即用代码片段覆盖 Claude Code、Cursor、Codex、Aider 等 10+ 工具 - ✅ **并排对比** — 即时对比各提供商的上下文窗口、速率限制和多模态能力 --- ## 三步上手 1. **选一个提供商** — 看下方 [Provider Directory](#provider-directory)。推荐从 **Groq** 开始（免信用卡，30 RPM 免费）。 2. **获取 API Key** — 点下方 [Quick Reference](#quick-reference--base-urls--api-keys) 里任意一个 [Get Key →] 链接，注册（大多只需邮箱），复制 key。不到 1 分钟。 3. **填入代码** — 复制 base URL + model ID，粘贴到下边的 [Quick Start](#quick-start--use-any-free-api-in-30-seconds) 示例。 配置具体工具？Claude Code · Cursor · Codex · OpenHuman · OpenCode · OpenClaw — 一键配置在 freellm.net/config/。 ## 快速上手 — 30 秒接入免费 API **没用过 API？** 最简单的方式：打开 console.groq.com/keys，用邮箱注册（无需信用卡），复制免费 key，粘贴到下面任意示例。不到一分钟就能跑起来。 以下所有提供商都暴露了 **OpenAI 兼容接口**。任何接受 `baseURL` + `apiKey` 的工具都能直接用 — 替换 base URL 和 key 即可。 ### Python (OpenAI SDK) ```python from openai import OpenAI client = OpenAI( base_url="https://api.groq.com/openai/v1", # 免费，无需信用卡 api_key="GROQ_API_KEY", # 获取: console.groq.com/keys ) response = client.chat.completions.create( model="llama-3.3-70b-versatile", # 模型列表见下方 Best Models 表 messages=[{"role": "user", "content": "你好！"}], ) print(response.choices[0].message.content) # Groq 免费额度: 30 RPM, 14,400 RPD — 个人使用绰绰有余 ``` ### Codex CLI ```bash export OPENAI_BASE_URL="https://api.groq.com/openai/v1" export OPENAI_API_KEY="your-groq-key" # 获取: console.groq.com/keys codex --model "llama-3.3-70b-versatile" ``` ### Cursor ``` Settings → Models → Add Model Model name: llama-3.3-70b-versatile Base URL: https://api.groq.com/openai/v1 API key: your-groq-key # 获取: console.groq.com/keys ``` ### Claude Code ```bash # Claude Code 需要 Anthropic 兼容接口 — 使用 OpenRouter export ANTHROPIC_BASE_URL="https://openrouter.ai/api" export ANTHROPIC_AUTH_TOKEN="sk-or-v1-your-key" # openrouter.ai/keys export ANTHROPIC_API_KEY="" # 必须为空 # 注意: OpenRouter 的 Anthropic 模型需充值 $10（一次性） ``` ### 其他工具？ 大多数 AI 开发工具都接受自定义 API 端点 — 拿到免费 key 后： - **Claude Code** — 设 `ANTHROPIC_BASE_URL` + `ANTHROPIC_AUTH_TOKEN`。详细配置 → - **Cursor** — Settings → Models → Add Model。详细配置 → - **Codex CLI** — 设 `OPENAI_BASE_URL` + `OPENAI_API_KEY`。详细配置 → - **OpenHuman** — 编辑 `config.toml`。详细配置 → - **Aider** — 编辑 `.aider.conf.yml`。详细配置 → - **Cline**（VS Code）— API provider 设置。详细配置 → - **Open WebUI** — Settings → Connections。详细配置 → 更多即用配置在 **freellm.net/config/**。 > **所有提供商的 Base URL 和 API Key 链接** 见下方 [Quick Reference](#quick-reference--base-urls--api-keys)。 --- ## 提供商目录 & 热门免费模型 | Provider | Free Models | Credit Card? | Max Context | Modalities | Get API Key | |---|---|---|---|---|---| | NVIDIA NIM | 123 | Phone verification | 1M | audio, embedding, image, reasoning, rerank, text, video, vision | → | | ModelScope | 55 | Registration | 1M | audio, image, reasoning, text, video, vision | → | | Cloudflare Workers AI | 39 | No | 10M | code, image, reasoning, text, video | → | | GitHub Models | 16 | No | 1M | image, pdf, reasoning, text | → | | Google Gemini | 15 | No | 1M | audio, image, pdf, reasoning, text, video, vision | → | | LLM7.io | 15 | No | 1M | audio, code, image, pdf, reasoning, text, video, vision | → | | OVHcloud AI Endpoints | 14 | Registration | 262K | audio, code, image, reasoning, text, video | → | | Ollama Cloud | 13 | Registration | 1M | code, image, reasoning, text, video, vision | → | | Groq | 12 | No | 262K | image, reasoning, text | → | | Mistral AI | 12 | No | 256K | code, image, text | → | | Cohere | 12 | No | 436K | image, text | → | | Kilo Code | 12 | No | 1M | audio, code, image, reasoning, text, video | → | | OpenCode Zen | 10 | Registration | 1M | audio, reasoning, vision | → | | Cerebras | 8 | No | 131K | image, reasoning, text | → | | Aion Labs | 7 | Registration | 131K | text | → | | Hugging Face | 7 | No | 131K | code, text | → | | Agnes AI | 5 | Registration | 256K | image, text, video, vision | → | | Alibaba Cloud Model Studio | 5 | Registration | 1M | code, image, text | → | | Z AI (Zhipu AI) | 4 | No | 200K | image, reasoning, text, video | → | | SambaNova | 4 | Registration | 128K | image, reasoning, text | → | | SiliconFlow | 3 | Registration | 131K | text | → | | xAI | 3 | Registration | 2M | text | → | | Chutes.ai | 2 | Registration | 131K | reasoning, text | → | | Glhf.chat | 2 | Registration | 131K | text | → | | Grok (xAI) | 2 | Registration | 131K | text | → | | AI21 Labs | 2 | Registration | 256K | text | → | | DeepSeek | 2 | Registration | 128K | text | → | | Nscale | 2 | Registration | 128K | text | → | | Nebius | 1 | Registration | 128K | text | → | | Provider | Free Models | Credit Model | Max Context | Modalities | Get API Key | |---|---|---|---|---|---| | OpenRouter | 24 | Free tier + $10 topup → 1K RPD | 1M | audio, code, embeddings, image, reasoning, rerank, speech, text, video | → | ## 快速参考 — Base URL 与 API Key | Provider | Base URL | Get API Key | Credit Card? | |---|---|---|---| | NVIDIA NIM | `https://integrate.api.nvidia.com/v1` | Get Key → | Phone verification | | ModelScope | `https://api-inference.modelscope.cn/v1` | Get Key → | Registration | | Cloudflare Workers AI | `https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run` | Get Key → | No | | OpenRouter | `https://openrouter.ai/api/v1` | Get Key → | Registration | | GitHub Models | `https://models.github.ai/inference` | Get Key → | No | | Google Gemini | `https://generativelanguage.googleapis.com/v1beta` | Get Key → | No | | LLM7.io | `https://api.llm7.io/v1` | Get Key → | No | | OVHcloud AI Endpoints | `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1` | Get Key → | Registration | | Ollama Cloud | `https://api.ollama.com` | Get Key → | Registration | | Groq | `https://api.groq.com/openai/v1` | Get Key → | No | | Mistral AI | `https://api.mistral.ai/v1` | Get Key → | No | | Cohere | `https://api.cohere.com/v2` | Get Key → | No | | Kilo Code | `https://api.kilo.ai/api/gateway` | Get Key → | No | | OpenCode Zen | `https://opencode.ai/zen/v1` | Get Key → | Registration | | Cerebras | `https://api.cerebras.ai/v1` | Get Key → | No | | Aion Labs | `https://api.aionlabs.ai/v1` | Get Key → | Registration | | Hugging Face | `https://router.huggingface.co/v1` | Get Key → | No | | Agnes AI | `https://apihub.agnes-ai.com/v1` | Get Key → | Registration | | Alibaba Cloud Model Studio | `https://dashscope-intl.aliyuncs.com/compatible-mode/v1` | Get Key → | Registration | | Z AI (Zhipu AI) | `https://open.bigmodel.cn/api/paas/v4` | Get Key → | No | | SambaNova | `https://api.sambanova.ai/v1` | Get Key → | Registration | | SiliconFlow | `https://api.siliconflow.cn/v1` | Get Key → | Registration | | xAI | `https://api.x.ai/v1` | Get Key → | Registration | | Chutes.ai | `https://api.chutes.ai/v1` | Get Key → | Registration | | Glhf.chat | `https://glhf.chat/api/openai/v1` | Get Key → | Registration | | Grok (xAI) | `https://api.x.ai/v1` | Get Key → | Registration | | AI21 Labs | `https://api.ai21.com/studio/v1` | Get Key → | Registration | | DeepSeek | `https://api.deepseek.com/v1` | Get Key → | Registration | | Nscale | `https://inference.api.nscale.com/v1` | Get Key → | Registration | | Nebius | `https://api.studio.nebius.com/v1` | Get Key → | Registration | ## 各提供商最佳免费模型 | Provider | Best Free Model | Model ID | Max Context | Rate Limit | |---|---|---|---|---| | NVIDIA NIM | z-ai/glm-5.2 | `z-ai/glm-5.2` | 1M | Up to 40 RPM | | | poolside/laguna-xs-2.1 | `poolside/laguna-xs-2.1` | 262K | Up to 40 RPM | | | z-ai/glm-5.1 | `z-ai/glm-5.1` | 202K | Up to 40 RPM | | ModelScope | MiniMax-M2.5-highspeed | `MiniMax/MiniMax-M2.5` | 204K | See provider | | | Qwen/Qwen3.5-35B-A3B | `qwen-qwen3-5-35b-a3b` | 131K | 2,000 RPD total; <=500 .. | | | Qwen/Qwen3.5-27B | `qwen-qwen3-5-27b` | 131K | 2,000 RPD total; <=500 .. | | Cloudflare Workers AI | Mistral 7B | `@cf/mistral/mistral-7b-instruct-v0.1` | 32K | See provider | | | Qwen 1.5 7B | `@cf/qwen/qwen1.5-7b-chat` | 32K | See provider | | | @cf/meta/llama-3.3-70b-instruct-fp8-fast | `@cf/meta/llama-3.3-70b-instruct-fp8-fast` | 131K | 10K neurons/day (shared) | | OpenRouter | NVIDIA: Nemotron 3 Ultra (free) | `nvidia/nemotron-3-ultra-550b-a55b:free` | 1M | See provider | | | Poolside: Laguna M.1 (free) | `poolside/laguna-m.1:free` | 262K | See provider | | | NVIDIA: Nemotron 3 Super (free) | `nvidia/nemotron-3-super-120b-a12b:free` | 262K | See provider | | GitHub Models | Phi-4 | `Phi-4` | 131K | See provider | | | Mistral Large (24.11) | `Mistral-large-2411` | 131K | See provider | | | AI21 Jamba 1.5 Large | `AI21-Jamba-1.5-Large` | 256K | See provider | | Google Gemini | Gemini 3.6 Flash | `gemini-3.6-flash` | 1M | 15 RPM, 1,500 RPD | | | Gemini 3.5 Flash | `gemini-3.5-flash` | 1M | 15 RPM, 1,500 RPD | | | Gemini 3.5 Flash-Lite | `gemini-3.5-flash-lite` | 1M | 30 RPM, 1,500 RPD | | LLM7.io | deepseek-r1-0528 | `deepseek-r1-0528` | 131K | 30 RPM (120 with token) | | | deepseek-v3-0324 | `deepseek-v3-0324` | 131K | 30 RPM (120 with token) | | | gpt-4o-mini | `gpt-4o-mini` | 131K | 30 RPM (120 with token) | | OVHcloud AI Endpoints | Qwen3.5-397B-A17B | `qwen3.5-397b-a17b` | 131K | 2 RPM (anonymous) | | | Meta-Llama-3_3-70B-Instruct | `meta-llama-3_3-70b-instruct` | 131K | 2 RPM (anonymous) | | | Qwen3.6-27B | `qwen3.6-27b` | 131K | 2 RPM (anonymous) | | Ollama Cloud | minimax-m3 | `minimax-m3` | 1M | Session/weekly limits (.. | | | gpt-oss:20b | `gpt-oss:20b` | 131K | Session/weekly limits (.. | | | nemotron-3-ultra | `nemotron-3-ultra` | 262K | Session/weekly limits (.. | | Groq | Moonshot Kimi K2 | `moonshotai/kimi-k2-instruct` | 131K | See provider | | | Moonshot Kimi K2 0905 | `moonshotai/kimi-k2-instruct-0905` | 131K | See provider | | | groq/compound | `groq/compound` | 131K | 30 RPM, 250 RPD | | Mistral AI | Mistral 7B | `open-mistral-7b` | 32K | See provider | | | Mixtral 8x7B | `open-mixtral-8x7b` | 32K | See provider | | | Mistral Medium 3.5 (128B) | `mistral-medium-3-5-128b` | 256K | ~1 RPS, 500K TPM | | Cohere | Command A+ (218B) | `command-a-218b` | 436K | 20 RPM | | | Command A (111B) | `command-a-111b` | 288K | 20 RPM | | | Command R+ | `command-r` | 128K | 20 RPM | | Kilo Code | nvidia/nemotron-3-ultra-550b-a55b:free | `nvidia/nemotron-3-ultra-550b-a55b:free` | 1M | ~200 req/hr | | | stepfun/step-3.7-flash:free | `stepfun/step-3.7-flash:free` | 262K | ~200 req/hr | | | nvidia/nemotron-3-super-120b-a12b:free | `nvidia/nemotron-3-super-120b-a12b:free` | 262K | ~200 req/hr | | OpenCode Zen | big-pickle | `big-pickle` | 0 | | | | DeepSeek V4 Flash | `deepseek-v4-flash-free` | 1M | | | | MiMo-V2.5 | `mimo-v2.5-free` | 1M | | | Cerebras | Llama 3.1 70B | `llama3.1-70b` | 131K | See provider | | | gpt-oss-120b | `gpt-oss-120b` | 131K | 5 RPM, 30K TPM, 1M TPD | | | zai-glm-4.7 (deprecated Aug 2026) | `zai-glm-4.7` | 131K | 5 RPM, 30K TPM, 1M TPD | | Aion Labs | Aion 2.5 | `aion-2-5` | 128K | 15 RPM, 20K TPD | | | Aion 2.0 | `aion-2-0` | 128K | 15 RPM, 20K TPD | | | Aion-RP 1.0 (8B) | `aion-rp-1-0-8b` | 32K | 15 RPM, 20K TPD | | Hugging Face | Meta-Llama-3.1-8B-Instruct | `meta-llama-3-1-8b-instruct` | 128K | Credit-metered | | | gemma-3-4b-it | `gemma-3-4b-it` | 131K | Credit-metered | | | Qwen2.5-Coder-7B-Instruct | `qwen2-5-coder-7b-instruct` | 131K | Credit-metered | | Agnes AI | agnes-1.5-flash | `agnes-1.5-flash` | 256K | 30 RPM | | | agnes-2.0-flash | `agnes-2.0-flash` | 256K | 30 RPM | | | agnes-image-2.0-flash | `agnes-image-2.0-flash` | 4K | 30 RPM (1K) | | Alibaba Cloud Model Studio | Qwen3-Max | `qwen3-max` | 128K | Tiered by region | | | Qwen3-Plus | `qwen3-plus` | 1M | Tiered by region | | | Qwen3-VL-Plus | `qwen3-vl-plus` | 128K | Tiered by region | | Z AI (Zhipu AI) | GLM-4.7-Flash | `glm-4.7` | 200K | 1 concurrent request | | | GLM-4.5-Flash | `glm-4.5` | 128K | 1 concurrent request | | | GLM-4.6V-Flash | `glm-4.6` | 128K | 1 concurrent request | | SambaNova | DeepSeek-V3.1 | `deepseek-v3-1` | 128K | 20 RPM, 20 RPD, 200K TPD | | | DeepSeek-V3.2 (Preview) | `deepseek-v3-2-preview` | 128K | 20 RPM, 20 RPD, 200K TPD | | | MiniMax-M2.7 | `minimax-m2-7` | 128K | 20 RPM, 20 RPD, 200K TPD | | SiliconFlow | deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | `deepseek-ai-deepseek-r1-distill-qwen-7b` | 131K | 30 RPM, 60K TPM | | | Abbreviation | `abbreviation` | 131K | See provider | | | deepseek-ai/DeepSeek-OCR | `deepseek-ai-deepseek-ocr` | 131K | 30 RPM, 60K TPM | | xAI | grok-4.3 | `grok-4-3` | 1M | Credit-based | | | grok-4.1-fast | `grok-4-1-fast` | 2M | Credit-based | | | grok-3-mini | `grok-3-mini` | 131K | Credit-based | | Chutes.ai | DeepSeek-R1 | `deepseek-ai/DeepSeek-R1` | 131K | Community-powered, no h.. | | | Llama 3.1 70B | `meta-llama/Meta-Llama-3.1-70B-Instruct` | 131K | Community-powered, no h.. | | Glhf.chat | Llama 3.1 70B | `meta-llama/Meta-Llama-3.1-70B-Instruct` | 131K | Unlimited for free models | | | Mixtral 8x7B | `mistralai/Mixtral-8x7B-Instruct-v0.1` | 32K | Unlimited for free models | | Grok (xAI) | Grok-2 | `grok-2` | 131K | $25/month free credits,.. | | | Grok-2 Mini | `grok-2-mini` | 131K | $25/month free credits,.. | | AI21 Labs | Jamba Large 1.7 | `jamba-large-1-7` | 256K | 200 RPM, 10 RPS | | | Jamba Mini 2 | `jamba-mini-2` | 256K | 200 RPM, 10 RPS | | DeepSeek | deepseek-chat (V3.2) | `deepseek-chat-v3-2` | 128K | Dynamic | | | deepseek-reasoner (R1) | `deepseek-reasoner-r1` | 128K | Dynamic | | Nscale | Llama-3.3-70B-Instruct | `llama-3-3-70b-instruct` | 128K | Fair-use | | | DeepSeek-R1-Distill-Llama-70B | `deepseek-r1-distill-llama-70b` | 128K | Fair-use | | Nebius | Qwen3-235B-A22B | `qwen3-235b-a22b` | 128K | Tier-based | ### 热门免费模型 | Model | Provider | Context | Weekly Usage | |---|---|---|---| | z-ai/glm-5.2 | NVIDIA NIM | 1M | 2998B tokens | | NVIDIA: Nemotron 3 Ultra (free) | OpenRouter | 1M | 2326B tokens | | Poolside: Laguna M.1 (free) | OpenRouter | 262K | 768B tokens | | NVIDIA: Nemotron 3 Super (free) | OpenRouter | 262K | 315B tokens | | Cohere: North Mini Code (free) | OpenRouter | 256K | 255B tokens | | poolside/laguna-xs-2.1 | NVIDIA NIM | 262K | 171B tokens | | z-ai/glm-5.1 | NVIDIA NIM | 202K | 158B tokens | | Poolside: Laguna S 2.1 (free) | OpenRouter | 262K | 83B tokens | | Poolside: Laguna XS 2.1 (free) | OpenRouter | 262K | 81B tokens | | NVIDIA: Nemotron 3 Nano 30B A3B (free) | OpenRouter | 256K | 45B tokens | --- ## 参与贡献 欢迎贡献！ - **补充缺失的免费模型** — 提交 issue 或 PR - **修正不准确的数据** — 速率限制会变、提供商会升级，欢迎 PR - **添加配置代码片段** — 有你常用工具的配置方案？添加到 `code-examples/` ### 收录标准 模型列入本列表需满足： 1. 提供商明确提供**免费层**（不仅仅是试用额度） 2. API **公开可访问**（无需候补、封闭内测、或逆向工程） 3. 试用额度：明确标注且最低 $1 额度 --- ## 链接 - 🌐 **在线网站**: freellm.net — 搜索、对比、在线体验、配置生成器 - 🔑 **API 密钥目录**: freellm.net/free-llm-api-keys/ - ⚙️ **配置生成器**: freellm.net/config/ - 🎮 **在线体验**: freellm.net/playground/ - 📊 **模型对比**: freellm.net/compare/ ## 许可证 MIT © open-free-llm-api ---数据每日自动更新 · 最后更新: 2026-08-12