Mistral
Efficient

Mistral Small 3.2

Efficient

Mistral's 24B small model: open-weights, fast, and strong on instruction following for its size.

Mistral Small 3.2 is a efficient AI model from Mistral. It costs $0.1000 per million input tokens and $0.3000 per million output tokens (blended $0.1600/M), with a 128,000-token context window.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$0.1000/M
per million input tokens
OUTPUT
$0.3000/M
per million output tokens
BLENDED 70/30
$0.1600/M
unchanged since 3 May
CONTEXT
128,000
tokens

Hand-maintained from Mistral's published pricing; no check date on record.

What it is good at
  • Open weights (Apache 2.0)
  • Single-GPU friendly
  • 128K context
Typical use cases
  • Self-hosted production chat
  • Cheap European-hosted inference

Benchmarks

vs. best public score
MMLU80.5%
Multitask academic knowledge across 57 subjects. Source: Mistral AI (self-reported), MMLU (model-card STEM table), read 2026-11-10.
LMArena Elo1356 Elo
Crowd-sourced head-to-head preference Elo rating. Source: LMArena (independent), LMArena Text Arena score, read 2026-11-10.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

Also reported by the provider

Results on benchmarks outside the set we compare across models. They are listed for completeness and are deliberately excluded from the bars above and from every ranking on this site, because a score is only comparable against the same benchmark variant. Why benchmark scores do not compare →

  • HumanEval Plus (Pass@5)92.9%self-reported by Mistral AI · 2025-06 · Model card STEM table lists only HumanEval Plus (Pass@5): 92.90% for Small 3.2 (88.99% for 3.1). Plain pass@1 is not published there. Not comparable with the HumanEval column, so shown here and never ranked.

More from Mistral

See all 15 →

Frequently asked questions

How much does Mistral Small 3.2 cost?

Mistral Small 3.2 costs $0.1000 per million input tokens and $0.3000 per million output tokens, for a blended reference rate of $0.1600 per million tokens.

What is Mistral Small 3.2's context window?

Mistral Small 3.2 supports up to 128,000 tokens of context in a single request.

What is Mistral Small 3.2 best for?

Mistral Small 3.2 is well suited to Open weights (Apache 2.0), Single-GPU friendly and 128K context.

Who makes Mistral Small 3.2?

Mistral Small 3.2 is developed and served by Mistral.

Terms used on this page