FreeLLM

One OpenAI-compatible endpoint. 15+ free LLM providers. No billing.

Aggregate the free tiers from Google, Groq, Cerebras, Mistral, OpenRouter, GitHub Models, Cohere, Cloudflare Workers AI, HuggingFace, Ollama, Pollinations, LLM7, OVH AI Endpoints — behind a single /v1 API. Smart routing, automatic failover, encrypted key storage, web dashboard. All features free.

Quick start

npx freellm

That's it. Opens a dashboard at http://localhost:3000.

What it does

1. Add your free API keys on the Keys page (Google, Groq, etc.)
2. Copy your unified key from the dashboard
3. Point any OpenAI client at http://localhost:3000/v1:

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:3000/v1",
    api_key="freellm-your-unified-key",
)

resp = client.chat.completions.create(
    model="auto",   # smart routing picks the best available provider
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)
print("Routed via:", resp.headers.get("x-routed-via"))

curl http://localhost:3000/v1/chat/completions \
  -H "Authorization: Bearer freellm-your-unified-key" \
  -H "Content-Type: application/json" \
  -d '{"model": "auto", "messages": [{"role": "user", "content": "hi"}]}'

Supported providers (all free tiers)

| Provider | Free models | Key needed | |---|---|---| | Google Gemini | Gemini 2.5 Flash, 2.0 Flash, 1.5 Flash | Yes (aistudio.google.com) | | Groq | Llama 3.3 70B, 3.1 8B, Gemma 2 9B | Yes (console.groq.com) | | Cerebras | Llama 3.3 70B, 3.1 8B | Yes (cloud.cerebras.ai) | | Mistral | Mistral Small, Nemo, Codestral | Yes (console.mistral.ai) | | OpenRouter | 20+ free models (:free suffix) | Yes (openrouter.ai) | | GitHub Models | GPT-4o, GPT-4o Mini | Yes (github.com) | | Cohere | Command R+, Command R | Yes (dashboard.cohere.com) | | Cloudflare | Llama 3.1 8B, Gemma 7B | Yes (ACCOUNT_ID:TOKEN) | | HuggingFace | Llama 3 8B, Mistral 7B | Yes (huggingface.co) | | Ollama | Any local model | No | | Pollinations | GPT-OSS, Mistral | No | | LLM7 | GPT-4o Mini, Llama 70B | No | | OVH AI Endpoints | Llama 3.3 70B, Mistral 7B | No |

Features

• OpenAI-compatible — /v1/chat/completions, /v1/models, streaming, tool calling, vision
• Smart routing — picks highest-priority healthy model for each request
• Automatic failover — on 429/5xx, skips to next provider in chain (up to 20 attempts)
• Rate limit tracking — in-memory per-provider counters, no key is sent over its limits
• Encrypted storage — API keys are AES-256-GCM encrypted at rest in SQLite
• Web dashboard — manage keys, reorder fallback chain, playground, analytics
• No billing — all features free, forever

Options

npx freellm --port 8080      # custom port (default: 3000)
npx freellm --no-open        # don't open browser automatically

Data storage

All data (database + encryption key) is stored in ~/.freellm/.

Requirements

• Node.js 22.5+ (uses built-in node:sqlite)

License

MIT