OpenRouter

One endpoint for hundreds of models. The openrouter/free alias costs $0 in and $0 out.

OpenRouter is a router, not a lab. One key, one /api/v1 URL, and it fans the request out to whoever actually serves the model. Anything whose id ends in :free is $0 in and $0 out; the same model without the suffix bills you. The openrouter/free alias picks a free model that matches the request.

The free pool is 18 models right now — Nemotron 3 Ultra, MiniMax M3, Laguna and GLM 5.2 among them — with up to 1M of context. DeepSeek and Qwen no longer have a :free endpoint. New accounts get 20 requests a minute and 50 a day; a $10 top-up lifts the daily cap to 1,000. Set the HTTP-Referer and X-Title headers if you want to show up on their public leaderboards.

Free models: 18
Max context: 1M
Free tier: 18 models at zero cost
Requirement: Email account, no card

Endpoint

Base URLhttps://openrouter.ai/api/v1

Env varOPENROUTER_API_KEY

Rate limit20 requests/minute, 50 requests/day; a $10 top-up raises the day to 1,000

OpenAI-compatible
text
image
audio
video
speech
embeddings
rerank
reasoning

Code examples

npm install openai

import OpenAI from 'openai'

const client = new OpenAI({
  baseURL: 'https://openrouter.ai/api/v1',
  apiKey: process.env.OPENROUTER_API_KEY
})

const response = await client.chat.completions.create({
  model: 'openrouter/free',
  messages: [{ role: 'user', content: 'Explain closures in one paragraph.' }]
})

console.log(response.choices[0].message.content)

pip install openai

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ["OPENROUTER_API_KEY"],
)

response = client.chat.completions.create(
    model="openrouter/free",
    messages=[{"role": "user", "content": "Explain closures in one paragraph."}],
)

print(response.choices[0].message.content)

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -d '{
    "model": "openrouter/free",
    "messages": [{"role": "user", "content": "Explain closures in one paragraph."}]
  }'

Worth knowing

Model ids ending in :free are the zero-cost ones; the same model without the suffix bills you.
Set the HTTP-Referer and X-Title headers to appear on the public leaderboards.

Models you get for free19

Ling 3.0 Flash Fininclusionai/ling-3.0-flash-fin:free
Dots 3 Note Previewdots-studio/dots-3-note-preview:free
LFM 2.5 2.6Bliquid/lfm-2.5-2.6b:free
Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning:free
Inkling Smallthinkingmachines/inkling-small:free
Laguna S 2.1poolside/laguna-s-2.1:free
Inklingthinkingmachines/inkling:free
Laguna XS 2.1poolside/laguna-xs-2.1:free
North Mini Codecohere/north-mini-code:free
GLM 5.2z-ai/glm-5.2:free
Nemotron 3.5 Content Safetynvidia/nemotron-3.5-content-safety:free
Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b:free
MiniMax M3minimax/minimax-m3:free
Nemotron 3 Nano Omninvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free
Gemma 4 26B A4Bgoogle/gemma-4-26b-a4b-it:free
Gemma 4 31Bgoogle/gemma-4-31b-it:free
MiniMax M2.7minimax/minimax-m2.7:free
Nemotron 3 Supernvidia/nemotron-3-super-120b-a12b:free
Free Models Routeropenrouter/free