Free-LLM — A Directory of Free AI/LLM APIs

What it is: A community-maintained, open-source directory that lists providers offering free access to large language models (LLMs). It covers 270+ free models from 51 providers, plus 9 local/self-hosted tools for unlimited private use. The project is backed by a live website at free-llm.com.

Why it exists

Finding a free LLM API usually means digging through changelogs, signing up for multiple platforms just to compare rate limits, or guessing which provider still has a free tier. This repo solves that by providing a structured, up-to-date reference.

Key features

• Community-maintained — users can vote, submit, and suggest edits; changes are moderated before publishing.
• Credit card transparency — every provider is labeled with whether it requires a credit card, phone verification, or nothing at all.
• Ready-to-run code — Python, JavaScript, and curl snippets for all 33 providers, plus configuration guides for coding tools like Claude Code, Cursor, Codex, and OpenCode.
• Side-by-side comparison — the website lets you compare two providers on limits, models, and pricing.

How to use it (3 steps)

1. Pick a provider — browse the directory. For beginners, the README suggests starting with Groq (no credit card, 30 requests per minute, 14,400 requests per day, free forever).
2. Get an API key — each provider row links directly to its key page. Most only need an email address.
3. Plug it in — copy the base URL and model ID into the provided code snippets. Most providers expose an OpenAI-compatible endpoint, so any tool that accepts a baseURL + apiKey works.

Provider categories

The directory splits providers into four groups:

• Permanent Free Tiers — ongoing free access with rate limits that never expire (e.g., Google AI Studio, Groq, Mistral, Cohere).
• Renewable Credits — free access that resets periodically (e.g., OpenRouter, Vercel AI Gateway).
• One-Time Trial Credits — sign up and get credits to use until depleted (e.g., Replicate, Cerebras, AI21 Labs).
• Local / Self-Hosted — tools you run on your own machine for unlimited, private, free use (e.g., Ollama, LM Studio, llama.cpp).

Quick start example

The README shows a Python snippet using the OpenAI SDK with Groq:

from openai import OpenAI

client = OpenAI(
    base_url="https://api.groq.com/openai/v1",  # free, no credit card
    api_key="GROQ_API_KEY",
)

response = client.chat.completions.create(
    model="llama-3.3-70b-versatile",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Additional resources

• Quick Reference table — lists base URLs and API key pages for every provider.
• Code examples folder — ready-to-copy snippets for all providers.
• Guides, Hall of Fame, and Submit pages — available on the website.
• License: MIT.
• Translations: English, Simplified Chinese, Traditional Chinese, Japanese, Korean.

Caveats to note

• Some entries include notes like "verify" or "not published" — the README is honest about uncertainty (e.g., payment method requirements, exact rate limits).
• Free tiers can change; the project relies on community updates to stay current.
• The README mentions a GitHub user report (dated 14 Sep 2026) about Nous Portal requiring a payment method — this is a user-reported observation, not a confirmed fact.

In short: Free-LLM is a one-stop reference for anyone who wants to use LLMs without paying, with clear labeling of requirements and ready-to-use code.