free-llm-apis
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseFree LLM API Setup
免费 LLM API 设置
Help users pick a free LLM provider and configure their API key. Every provider here has a permanent free tier, no credit card needed.
帮助用户选择免费的 LLM 提供商并配置其 API key。此处列出的每个提供商都有永久免费额度，无需信用卡。
Provider Selection
提供商选择
Ask the user what matters most, then recommend accordingly:
| Priority | Best picks |
|---|---|
| Highest rate limits | Groq (30 RPM, 14.4K RPD), Cerebras (30 RPM, 14.4K RPD) |
| Largest model selection | Cloudflare Workers AI (49+ models), OpenRouter (32+ models) |
| Strongest proprietary models | Google Gemini (Gemini 2.5 Pro), GitHub Models (GPT-4o) |
| Fastest inference | Groq, Cerebras (both optimized for speed) |
| Highest token budget | Mistral AI (1B tokens/month) |
| European provider | Mistral AI (EU), LLM7.io (UK) |
| No signup required | LLM7.io (basic tier works without token) |
先询问用户最看重哪些因素，再据此推荐：
| 优先事项 | 推荐选择 |
|---|---|
| 最高速率限制 | Groq（30 RPM，14.4K RPD）、Cerebras（30 RPM，14.4K RPD） |
| 模型数量最多 | Cloudflare Workers AI（49+ 个模型）、OpenRouter（32+ 个模型） |
| 最强的专有模型 | Google Gemini（Gemini 2.5 Pro）、GitHub Models（GPT-4o） |
| 最快推理 | Groq、Cerebras（两者均针对速度进行了优化） |
| 最高 token 配额 | Mistral AI（每月 1B tokens） |
| 欧洲提供商 | Mistral AI（EU）、LLM7.io（UK） |
| 无需注册 | LLM7.io（基础层级无需 token 即可使用） |
Provider categories
提供商类别
Provider APIs -- run by the companies that train the models:
- Google Gemini, Cohere, Mistral AI, Zhipu AI
- See references/provider-apis.md for setup instructions.
Inference providers -- third-party platforms hosting open-weight models:
- GitHub Models, NVIDIA NIM, Groq, Cerebras, Cloudflare Workers AI, LLM7.io, Kluster AI, OpenRouter, Hugging Face
- See references/inference-providers.md for setup instructions.
提供商 API——由训练模型的公司运营：
- Google Gemini、Cohere、Mistral AI、Zhipu AI
- 有关设置说明，请参阅 references/provider-apis.md。
推理提供商——托管 open-weight 模型的第三方平台：
- GitHub Models、NVIDIA NIM、Groq、Cerebras、Cloudflare Workers AI、LLM7.io、Kluster AI、OpenRouter、Hugging Face
- 有关设置说明，请参阅 references/inference-providers.md。
Workflow
工作流程
- Ask what models, rate limits, or features the user cares about. If they already know which provider they want, skip to step 3.
- Match their priorities against the table above. Suggest 1-2 options with a short reason.
- Load the right reference file and walk through the setup steps for that provider (API key, code example, env var).
- Offer a quick test script or curl command so they can confirm the key works.
- 询问用户最关注哪些模型、速率限制或功能。如果用户已经知道想使用哪个提供商，则直接跳到步骤 3。
- 根据上表匹配用户的优先事项。推荐 1-2 个选项，并简要说明原因。
- 加载相应的参考文件，并引导用户完成该提供商的设置步骤（API key、代码示例、环境变量）。
- 提供一个快速测试脚本或 curl 命令，方便用户确认 key 是否可用。
Quick Test Template
快速测试模板
After setup, use this to verify any provider:
python
from openai import OpenAI
client = OpenAI(api_key="KEY", base_url="BASE_URL")
response = client.chat.completions.create(
model="MODEL_NAME",
messages=[{"role": "user", "content": "Say hello in one sentence."}],
max_tokens=50
)
print(response.choices[0].message.content)Replace KEY, BASE_URL, and MODEL_NAME with values from the provider's setup guide.
设置完成后，使用以下代码验证任意提供商：
python
from openai import OpenAI
client = OpenAI(api_key="KEY", base_url="BASE_URL")
response = client.chat.completions.create(
model="MODEL_NAME",
messages=[{"role": "user", "content": "Say hello in one sentence."}],
max_tokens=50
)
print(response.choices[0].message.content)将 KEY、BASE_URL 和 MODEL_NAME 替换为提供商设置指南中的对应值。
Key Notes
重要说明
- All endpoints work with the OpenAI SDK unless noted.
- RPM = requests per minute. RPD = requests per day.
- Google Gemini's free tier is blocked in the EEA, UK, and Switzerland.
- "Limits undocumented" = the provider doesn't publish rate limits.
- Don't hardcode API keys. Use environment variables or a secrets manager.
- 除非另有说明，所有端点均可与 OpenAI SDK 配合使用。
- RPM = 每分钟请求数。RPD = 每天请求数。
- Google Gemini 的免费层级在 EEA、英国和瑞士不可用。
- "Limits undocumented" = 提供商未公布速率限制。
- 不要硬编码 API key。请使用环境变量或 secrets manager。