Mistral AI offers frontier open and commercial models via REST API. Mistral Large 2 rivals top closed models; Codestral specializes in code generation; Mistral Embed handles semantic search. The API is OpenAI-compatible — simply change the base URL and model name.
https://api.mistral.ai/v1
Auth type
Bearer Token
Auth header
Authorization: Bearer ...
Rate limit
1 request/s · 500,000 TPM
Pricing
from $0.10/mo
Free quota
Experimental models free during preview
Documentation
https://docs.mistral.ai
Endpoint status
Server online — HTTP 404 — server is online but path returned an error (may require auth)259ms
(checked Mar 29, 2026)
Builder score
C
63%
builder-friendly
Standard Bearer token authentication using your Mistral API key.
Authorization: Bearer ...
Mistral Medium 3: $0.40/M input, $2.00/M output. Mistral Small 3.1: $0.03/M input, $0.11/M output. Mistral Nemo: $0.02/M. Codestral: $0.20/M input, $0.60/M output. Mistral Large 2: $2.00/M input, $6.00/M output. Free experiment plan available.
| Method | Path | Description |
|---|---|---|
| POST | /chat/completions |
OpenAI-compatible chat completions |
| POST | /embeddings |
Generate text embeddings |
| POST | /fim/completions |
Fill-in-the-middle for code (Codestral) |
| GET | /models |
List available models |
curl https://api.mistral.ai/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $MISTRAL_API_KEY" \
-d '{
"model": "mistral-large-latest",
"messages": [{"role": "user", "content": "Write a haiku about APIs."}]
}'
{
"id": "cmpl-e5cc70bb28c444948073e77776eb30ef",
"object": "chat.completion",
"model": "mistral-large-latest",
"choices": [{
"index": 0,
"message": {
"role": "assistant",
"content": "Keys unlock the gate / Data flows through silent paths / Endpoints never sleep"
},
"finish_reason": "stop"
}],
"usage": {"prompt_tokens": 15, "completion_tokens": 18}
}
Data sourced from API Map. Always verify pricing and rate limits against the official Mistral AI documentation.