Language modelDeprecated

Mistral Small 3.2

Mistral AI

Released: June 20, 2025
Data date: September 8, 2026
Predecessor: Mistral Small 3.1
Successor: Mistral Small 4

Speed: 156.2 tokens/s
Output throughput
Price: No current rate
Historical rates under Costs
Input: Text, Image
Output: Text

New in Mistral Small 3.2 are more precise instruction following, fewer endless or repetitive generations, and more robust function calls compared with Mistral Small 3.1. It accepts text and images with 128,000 tokens of context and uses Apache 2.0.

Mistral’s current documentation lists Small 3.2 as deprecated and points to Mistral Small 4.

Specifications and access

Specification	Value and source
Weights	Open weightsSource
Access	LocalSource
Parameters	24BSource
Context window	131,072 tokensSource
License	Apache-2.0Source
Release	2025-06-20Source
Parameters	24 BSource
Input	Text, ImageSource
Output	TextSource
API model ID	mistral-small-2506Source
Tool calling	YesSource
Source status	deprecatedSource

Measurements without matching peer values

These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.

WildBench (Show measurement, test conditions, and source)

Source value: 65.33
Score: 65.33
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card comparison table; exact judge/config not stated

Source Mistral Small 3.2 model card

Arena Hard (Show measurement, test conditions, and source)

Source value: 43.1
Score: 43.1
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card comparison table; exact judge/config not stated

Source Mistral Small 3.2 model card

IF (Show measurement, test conditions, and source)

Source value: 84.78
Score: 84.78
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card comparison table; exact metric setup not stated

Source Mistral Small 3.2 model card

Infinite Generations (Show measurement, test conditions, and source)

Source value: 1.29
Score: 1.29
Metric: percent
Unit: %
Direction: Lower is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card table; lower is better

Source Mistral Small 3.2 model card

MMLU (Show measurement, test conditions, and source)

Source value: 80.5
Score: 80.5
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

MMLU Pro (Show measurement, test conditions, and source)

Source value: 69.06
Score: 69.06
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

MATH (Show measurement, test conditions, and source)

Source value: 69.42
Score: 69.42
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

GPQA Main (Show measurement, test conditions, and source)

Source value: 44.22
Score: 44.22
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

GPQA Diamond (Show measurement, test conditions, and source)

Source value: 46.13
Score: 46.13
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

MBPP Plus (Show measurement, test conditions, and source)

Source value: 78.33
Score: 78.33
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

HumanEval Plus (Show measurement, test conditions, and source)

Source value: 92.9
Score: 92.9
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

SimpleQA (Show measurement, test conditions, and source)

Source value: 12.1
Score: 12.1
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Mistral Small 3.2 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruct table; Model-card benchmark table

Source Mistral Small 3.2 model card

Page 1 of 3

Predecessor and successor

Predecessor: Mistral Small 3.1
Model lineage evidence
Mistral's own docs describe Mistral Small 3.2 as 'an update to our previous small model, released June 2025', with Mistral Small 3.1 (March 2025) being the only prior Small release.
Mistral Docs, Mistral Small 3.2 model card (2026-09-19)
Later model generation: Mistral Small 4
Next generation in the directory
Model lineage evidence
Mistral's launch post benchmarks Small 4's throughput '3x more requests per second... compared to Mistral Small 3' and describes its no-reasoning mode as 'equivalent to the same chat style of Mistral Small 3.2', the most recent Small release in this directory.
Mistral AI, "Introducing Mistral Small 4" (2026-09-19)

Sources and data date

Type	Evidence and data date
Speed	Mistral. Output speed after the first streaming chunk, not total response time. Classified as deprecated by Artificial Analysis. Artificial Analysis maintains only the default 10K-input workload there. Retrieved September 8, 2026. Source: Artificial Analysis · Artificial Analysis (retrieved September 8, 2026)Speed measurement methodology
Input	Documented modalities. Retrieved September 8, 2026. Source: Model documentation · Model metadata (retrieved September 8, 2026) · Mistral Small 3.2 model card · Editorial description reviewed October 2, 2026
Output	Documented modalities. Retrieved September 8, 2026. Source: Model documentation · Model metadata (retrieved September 8, 2026) · Mistral Small 3.2 model card · Editorial description reviewed October 2, 2026
Research status	Research date September 8, 2026. 3 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not found in the inspected sources	knowledge-cutoff Original wording No exact cutoff in checked docs/card.
Not found in the inspected sources	training-tokens Original wording No token count in checked sources.
Not found in the inspected sources	max-output-tokens Original wording No model-specific maximum output found.
Additional source	Context window data
Additional source	API pricing
Additional source	Model metadata (retrieved September 8, 2026) · Mistral Small 3.2 model documentation · Editorial description reviewed October 2, 2026
Additional source	Mistral model overview · Editorial description reviewed October 2, 2026
Additional source	OpenRouter: Popularity index (derived), CC BY 4.0
Additional source	Vercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional source	Creative Commons Attribution 4.0
Additional source	Mistral AI, "Introducing Mistral Small 4" (2026-09-19)

Mistral Small 3.2

Specifications and access

Measurements without matching peer values

Our tests

Pelican on a bicycle

Popularity

Published prices

Predecessor and successor

Sources and data date