Skip to content
Get an API key

Live · serverless · 2 regions

Mistral Large 3

by Mistral AI

Mistral's largest open model, a mixture-of-experts design for long documents, tool use and European languages.

  • chat
  • vision
  • FP8
Serverless priceUSD
Input
$0.65per 1M tokens
Output
$1.60per 1M tokens

Billed per token, no minimums.

Reference

Specifications

Model ID
mistral-large-3
Endpoint
POST /v1/chat/completions
Publisher
Mistral AI
Tasks
chat · vision
Served context
256K tokens
Quantization
FP8
Decode speed
58 tok/smedian, one request
p50 first token
270 mslast 24 hours
Regions
us-east · eu-west
Dedicated
On request

Last 24 hours

Latency and speed

Median of each 15-minute step on the shared pool, measured at the API edge. Time to first token includes queueing; decode speed is for one request.

Time to first token · ms

270 msmedian

p95 288 ms

Decode speed · tok/s

58 tok/smedian

p5 54 tok/s

Show the data
Median of each 15-minute step over the last 24 hours, newest first
StepTime to first tokenDecode speed
15m ago287 ms58 tok/s
30m ago274 ms57 tok/s
45m ago270 ms57 tok/s
1h ago279 ms56 tok/s
1h 15m ago280 ms57 tok/s
1h 30m ago282 ms58 tok/s
1h 45m ago267 ms60 tok/s
2h ago275 ms57 tok/s
2h 15m ago287 ms57 tok/s
2h 30m ago281 ms58 tok/s
2h 45m ago268 ms60 tok/s
3h ago269 ms59 tok/s
3h 15m ago279 ms59 tok/s
3h 30m ago268 ms58 tok/s
3h 45m ago275 ms56 tok/s
4h ago287 ms58 tok/s
4h 15m ago277 ms56 tok/s
4h 30m ago274 ms58 tok/s
4h 45m ago285 ms57 tok/s
5h ago265 ms59 tok/s
5h 15m ago278 ms59 tok/s
5h 30m ago286 ms56 tok/s
5h 45m ago280 ms57 tok/s
6h ago288 ms54 tok/s
6h 15m ago285 ms57 tok/s
6h 30m ago264 ms60 tok/s
6h 45m ago287 ms55 tok/s
7h ago272 ms59 tok/s
7h 15m ago281 ms56 tok/s
7h 30m ago274 ms59 tok/s
7h 45m ago263 ms59 tok/s
8h ago278 ms56 tok/s
8h 15m ago268 ms58 tok/s
8h 30m ago261 ms57 tok/s
8h 45m ago283 ms56 tok/s
9h ago270 ms59 tok/s
9h 15m ago275 ms56 tok/s
9h 30m ago272 ms58 tok/s
9h 45m ago271 ms60 tok/s
10h ago270 ms58 tok/s
10h 15m ago258 ms59 tok/s
10h 30m ago276 ms58 tok/s
10h 45m ago261 ms58 tok/s
11h ago255 ms61 tok/s
11h 15m ago263 ms59 tok/s
11h 30m ago261 ms58 tok/s
11h 45m ago265 ms60 tok/s
12h ago255 ms61 tok/s
12h 15m ago345 ms51 tok/s
12h 30m ago272 ms57 tok/s
12h 45m ago255 ms59 tok/s
13h ago269 ms60 tok/s
13h 15m ago272 ms58 tok/s
13h 30m ago253 ms61 tok/s
13h 45m ago260 ms59 tok/s
14h ago261 ms60 tok/s
14h 15m ago246 ms60 tok/s
14h 30m ago260 ms59 tok/s
14h 45m ago248 ms62 tok/s
15h ago343 ms51 tok/s
15h 15m ago260 ms58 tok/s
15h 30m ago264 ms60 tok/s
15h 45m ago267 ms58 tok/s
16h ago268 ms57 tok/s
16h 15m ago247 ms62 tok/s
16h 30m ago270 ms58 tok/s
16h 45m ago247 ms62 tok/s
17h ago256 ms61 tok/s
17h 15m ago329 ms53 tok/s
17h 30m ago264 ms58 tok/s
17h 45m ago273 ms57 tok/s
18h ago265 ms60 tok/s
18h 15m ago267 ms60 tok/s
18h 30m ago270 ms56 tok/s
18h 45m ago271 ms57 tok/s
19h ago367 ms46 tok/s
19h 15m ago259 ms58 tok/s
19h 30m ago263 ms60 tok/s
19h 45m ago273 ms58 tok/s
20h ago259 ms58 tok/s
20h 15m ago272 ms57 tok/s
20h 30m ago258 ms61 tok/s
20h 45m ago276 ms56 tok/s
21h ago276 ms56 tok/s
21h 15m ago270 ms59 tok/s
21h 30m ago259 ms58 tok/s
21h 45m ago262 ms58 tok/s
22h ago270 ms60 tok/s
22h 15m ago272 ms59 tok/s
22h 30m ago263 ms60 tok/s
22h 45m ago275 ms59 tok/s
23h ago278 ms59 tok/s
23h 15m ago273 ms56 tok/s
23h 30m ago281 ms58 tok/s
23h 45m ago268 ms58 tok/s
24h ago267 ms60 tok/s

API

Make a request

The same request shape as every other model on /v1/chat/completions. Switching between them is a change to one string.

curl https://api.weir.example/v1/chat/completions \
  -H "Authorization: Bearer $WEIR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral-large-3",
    "messages": [
      { "role": "user", "content": "What does a KV cache do?" }
    ]
  }'
POST /v1/chat/completionsapi.weir.example

Dedicated

Need reserved capacity?

Mistral Large 3 runs on the shared serverless pool. Dedicated replicas for it are set up on request, with the same API and the same model ID.

Talk to an engineer

Compare

Similar models