Seminal AI
§1

Mistral Small 4

Best price/performance workhorse — a hybrid instruct+reasoning+coding model with only 6.5B active params, so it is fast and cheap while still handling agents and vision; the model to self-host under Apache 2.0 if you want reasoning without a licence conversation.

Data checked 2026-09-06

Mistral AI ga open weights mistral-small-2603

Context window
256K tokens in
Max output
Input
$0.15 per 1M tokens
Output
$0.60 per 1M tokens
Cached input
$0.01 per 1M tokens
Blended
$0.26 3:1 in:out
Released
2026-03-16
Parameters
119B-A6.5B MoE
Licence
Apache-2.0

A What it is for

Alias: mistral-small-latest.

At a three-to-one input-to-output ratio, Mistral Small 4 costs $0.26 per million tokens blended. A workload of one million input and 330,000 output tokens per day would run about $10.44 per month at list price, before caching or batch discounts.

B Capabilities

tools vision reasoning prompt-caching batch streaming structured-output web-search
Input modalitiestext, image, pdf
Output modalitiestext
Batch discount50% off asynchronous requests

C Others from Mistral AI

Model Context Max out In $/M Out $/M Blended Status
Mistral Medium 3.5 mistral-medium-3-5 256K $1.50 $7.50 $3.00 ga open
Mistral Large 3 mistral-large-2512 256K $0.50 $1.50 $0.75 ga open
Ministral 3 14B ministral-14b-2512 256K $0.20 $0.20 $0.20 ga open
Ministral 3 8B ministral-8b-2512 256K $0.15 $0.15 $0.15 ga open
Ministral 3 3B ministral-3b-2512 256K $0.10 $0.10 $0.10 ga open
Z.ai GLM 5.2 zai-glm-5-2 1M 128K $1.40 $4.40 $2.15 preview open
Codestral codestral-2508 128K $0.30 $0.90 $0.45 ga
Leanstral 1.5 labs-leanstral-1-5 256K 128K $0.00 $0.00 $0.00 preview open
OCR 4.1 mistral-ocr-4-1 ga
OCR 4.0 mistral-ocr-4-0 ga
OCR 3 mistral-ocr-2512 ga
Voxtral Small voxtral-small-2507 32K $0.10 $0.40 $0.18 ga open

D Verify before you commit

Model pricing changes without notice and this page is a snapshot. Confirm against the vendor's own page before you build a budget on it.

Source: docs.mistral.ai · checked 2026-09-06 · Mistral AI pricing · API docs