Skip to content

Getting started

1. Install

pip install freetheai

Python 3.9 or newer. The only dependency is httpx.

2. Get a key

  1. Sign up at freetheai.org.
  2. Open API keys and create a key. It starts with ftai_ and is shown once, so copy all of it.
  3. Free models need the daily check-in, one quick captcha a day.

The free tier is 50 requests a day, reset at 00:00 UTC. The plan page shows how to raise it.

Put the key in the environment:

export FREETHEAI_API_KEY="ftai_..."

$env:FREETHEAI_API_KEY = "ftai_..."

3. Pick a model

Model IDs look like fta/<provider>/<model>. Browse them on freetheai.org/models or list them in code:

from freetheai import FreeTheAI

client = FreeTheAI()
for model in client.models.list():
    print(model["id"])

4. Send a request

reply = client.chat.completions.create(
    model="fta/zai/glm-5.3",
    messages=[
        {"role": "system", "content": "You are a concise assistant."},
        {"role": "user", "content": "What is an API gateway?"},
    ],
)
print(reply["choices"][0]["message"]["content"])
print(reply["usage"])

Replies use the OpenAI Chat Completions shape, returned as plain dictionaries.

Already using the OpenAI SDK?

FreeTheAI speaks the OpenAI wire format, so the openai package works with a different base URL:

from openai import OpenAI

client = OpenAI(base_url="https://api.freetheai.org/v1", api_key="ftai_...")

That gives you the plain client without the handling described on Reliability.