codestral-latest

Served by Mistral (Codestral) through this free, OpenAI-compatible gateway. Pin it by name, or let the router pick it.

What we have measured

Not the advertised numbers. These are from this gateway's own request log over the last 30 days, recomputed on every page load.

Measured end to end from our server, including any failed attempts on other providers before this one answered, so it is what a caller experienced rather than what the upstream clocked.

Call it

curl https://dreamprompting.com/api/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "codestral/codestral-latest",
    "messages": [{"role":"user","content":"Hello!"}]
  }'

from openai import OpenAI

client = OpenAI(base_url="https://dreamprompting.com/api/v1",
                api_key="YOUR_API_KEY")

r = client.chat.completions.create(
    model="codestral/codestral-latest",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(r.choices[0].message.content)

Get a free key, or try it with no account in the playground. Full parameters are in the API docs.

What happens when it is busy

Free tiers run out. When Mistral (Codestral) rate limits this model, the router moves to the next provider in the chain and the request still answers, from a different model. Send "allow_fallback": false to refuse the substitution and get an error instead, which is what you want if your application announces which model it is using.

If you would rather not share the allowance at all, add your own Mistral (Codestral) key: requests then run on your account, with no daily quota. See pricing.