How to use
- Copy the installation command and review the source package before installing it in your own agent environment.
- Check the source document for required tools, credentials and supported hosts.
- Describe your input and desired deliverable; review the output against the source material.
npx skills add https://github.com/mnfst/awesome-free-llm-apis --skill free-llm-apisBefore you start
- Installation does not include a model subscription, API allowance or permission to access your files.
- Repository stars and installation counts describe activity, not an independent quality or security test.
Original skill document
Rendered from the source repository. Headings, examples, code, tables, links, and referenced images are preserved.
Free LLM API Setup
Help users pick a free LLM provider and configure their API key. Every provider here has a permanent free tier, no credit card needed.
Provider Selection
Ask the user what matters most, then recommend accordingly:
| Priority | Best picks |
|---|---|
| Fastest inference | Groq (LPU hardware; ~1,000 RPD on chat models) |
| Largest model selection | NVIDIA NIM (100+ models), Cloudflare Workers AI (60+), Hugging Face (thousands, small credit) |
| Strongest proprietary models | Google Gemini, Cohere (non-commercial only) |
| Free frontier open-weight models | OpenRouter (:free models), Kilo Code, NVIDIA NIM |
| European provider / EU hosting | OVHcloud AI Endpoints (FR, EU data centers), LLM7.io (UK) |
| No signup or key required | OVHcloud AI Endpoints (2 RPM/IP), Kilo Code (200 req/hr/IP), LLM7.io (anonymous tier) |
| Privacy-sensitive prompts | Avoid Gemini free tier outside the EEA, Switzerland and the UK, OpenRouter/Kilo free routes, NVIDIA trial endpoints |
Provider categories
<!-- providers:start --> Provider APIs -- run by the companies that train the models:
- Aion Labs, Cohere, Google Gemini, Z AI (Zhipu AI)
- See references/provider-apis.md for setup instructions.
Inference providers -- third-party platforms hosting open-weight models:
- AxForge, Cloudflare Workers AI, Gonka Broker, Groq, Kilo Code (no key needed), LLM7.io (no key needed), ModelScope, NVIDIA NIM, OpenRouter, OVHcloud AI Endpoints (no key needed), SiliconFlow
- See references/inference-providers.md for setup instructions.
<!-- providers:end -->
Workflow
- 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.
Quick Test Template
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, BASEURL, and MODELNAME with values from the provider's setup guide.
Key Notes
- All endpoints work with the OpenAI SDK unless noted.
- RPM = requests per minute. RPD = requests per day.
- Google Gemini's free tier works in the EEA, UK, and Switzerland, but its terms require paid services if you ship an app to users there.
- Free tiers, model lists, and limits change often. The reference files are generated from
data.json; check the provider's own docs before relying on a limit. - "Limits undocumented" = the provider doesn't publish rate limits.
- Don't hardcode API keys. Use environment variables or a secrets manager.