Early build · scores computed by the deterministic engine from real, sourced developments (30 days) · weights v1.0-prior
LatestModels › Mistral AI
Vikshy Models · Mistral AI

Mistral AI models: pricing, context & what each is for

Mistral AI has 12 models in the Vikshy catalogue, spanning Mistral 3, Magistral, Codestral, Devstral and more, compared side by side on context window, API pricing, modality, and what each is best for, every model linked to Mistral AI's official docs and pricing. Prices are per million tokens and can change, so the linked pricing page is the source of truth. For how Mistral AI ranks against other labs, see its Vikshy Score. Updated 2026-07-21.

ModelContextInput /MOutput /MModalityAccess
Mistral Medium 3.5Frontier
Complex agent pipelines and premium coding assistants.
256K
$1.50
$7.50
Multimodal
API
Mistral Large 3Flagship
Teams wanting a top open weight generalist with a hosted option.
256K
$0.50
$1.50
Multimodal
API + open weights
Magistral MediumReasoning
Math, logic, and multi step problem solving.
128K
Text
API
Magistral SmallReasoning
Local reasoning and cost controlled deployments.
128K
Open
Open
Text
API + open weights
Mistral Small 4Balanced
High throughput budget tasks that still need long context.
256K
Text
API + open weights
Devstral MediumBalanced
Autonomous coding agents and repo scale edits.
256K
$0.40
$2.00
Text
API
CodestralEfficient
IDE autocomplete and high frequency code generation.
256K
$0.30
$0.90
Text
API
Ministral 3 14BEfficient
On device and latency sensitive text tasks.
128K
Open
Open
Text
API + open weights
Voxtral MiniEfficient
Voice interfaces and transcription pipelines.
32K
Open
Open
Audio
API + open weights
Devstral SmallOpen
Self hosted coding agents on a budget.
256K
$0.10
$0.30
Text
API + open weights
Pixtral 12BOpen
Image plus text apps that want open weights.
128K
$0.15
$0.15
Multimodal
API + open weights
Mistral Large 2Legacy
Workloads pinned to the older Large line.
128K
$2.00
$6.00
Text
API + open weights

Mistral AI models, common questions

What is Mistral AI's most capable model?

Its highest-tier model in the catalogue is Mistral Medium 3.5 (Frontier tier), built for complex agent pipelines and premium coding assistants.

Which Mistral AI model is cheapest?

By API input price, Devstral Small is the lowest at $0.10 per million input tokens ($0.30 output).

What is the largest context window Mistral AI offers?

Mistral Medium 3.5 has the largest listed context window, up to 256K tokens.

Does Mistral AI offer open-weight models?

Yes. 8 of Mistral AI's catalogued models are available as open weights.