GeekRadar / The comparison desk

COMPARE.

AI models

mistral

Mistral Small 3.2

AI model Released Jun 20, 2025

Context window: 128,000 tokens
Maximum output: 16,384 tokens
Open weights: Yes

Capability & preference

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

Mistral Small 3.2 1338.0 mistral-small-2506 · 95% interval 1332.8–1343.2 · 17,462 votes · 2026-10-15

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

Mistral Small 3.2

Tested configuration Radar task score

Unrated

1/4 core tests · not enough evidence

2 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

Mistral Small 3.2 49.1^*% Approx. 95% interval: 43.6–54.5% · 2026-08-28

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

Mistral Small 3.2 Not available

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

Mistral Small 3.2 Not available

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

Mistral Small 3.2 Not available

Competition mathematics

OTIS Mock AIME 2024-2025 · Separate result

Competition problems with model-extracted answers checked against an exact key. Earlier grading procedures are excluded.

Mistral Small 3.2 30.3^*% Approx. 95% interval: 19.0–41.6% · 2026-08-30

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,340.39 rating

Confidence interval: 1,322.95 – 1,357.83

1093 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,361.99 rating

Confidence interval: 1,351.92 – 1,372.06

3337 votes

Oct 15, 2026

Category: coding
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,304.75 rating

Confidence interval: 1,292.04 – 1,317.46

2149 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,357.10 rating

Confidence interval: 1,350.14 – 1,364.05

7567 votes

Oct 15, 2026

Category: english
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,291.61 rating

Confidence interval: 1,284.37 – 1,298.85

12179 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,290.91 rating

Confidence interval: 1,271.49 – 1,310.34

856 votes

Oct 15, 2026

Category: expert
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,336.55 rating

Confidence interval: 1,297.47 – 1,375.64

227 votes

Oct 15, 2026

Category: french
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,340.18 rating

Confidence interval: 1,310.96 – 1,369.41

427 votes

Oct 15, 2026

Category: german
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,335.06 rating

Confidence interval: 1,328.17 – 1,341.96

7757 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,356.77 rating

Confidence interval: 1,346.97 – 1,366.57

3448 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,325.03 rating

Confidence interval: 1,314.10 – 1,335.96

2926 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,303.76 rating

Confidence interval: 1,293.55 – 1,313.97

3142 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,340.22 rating

Confidence interval: 1,323.23 – 1,357.22

1142 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,338.57 rating

Confidence interval: 1,327.65 – 1,349.50

2898 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,351.31 rating

Confidence interval: 1,333.20 – 1,369.43

994 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,345.76 rating

Confidence interval: 1,328.15 – 1,363.37

1128 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,360.33 rating

Confidence interval: 1,352.61 – 1,368.05

5904 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,310.02 rating

Confidence interval: 1,300.40 – 1,319.64

3757 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,309.20 rating

Confidence interval: 1,300.35 – 1,318.05

4383 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,275.28 rating

Confidence interval: 1,249.33 – 1,301.23

520 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,257.69 rating

Confidence interval: 1,226.09 – 1,289.30

359 votes

Oct 15, 2026

Category: korean
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,325.88 rating

Confidence interval: 1,315.90 – 1,335.87

3421 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,341.09 rating

Confidence interval: 1,323.71 – 1,358.46

1049 votes

Oct 15, 2026

Category: math
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,343.29 rating

Confidence interval: 1,332.60 – 1,353.98

2995 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,315.08 rating

Confidence interval: 1,308.75 – 1,321.40

9885 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,338.02 rating

Confidence interval: 1,332.83 – 1,343.20

17462 votes

Oct 15, 2026

Category: overall
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / polishhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,316.76 rating

Confidence interval: 1,302.14 – 1,331.39

1736 votes

Oct 15, 2026

Category: polish
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,323.05 rating

Confidence interval: 1,307.04 – 1,339.05

1232 votes

Oct 15, 2026

Category: russian
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-15 · rating

Mistral Small 3.2

1,346.94 rating

Confidence interval: 1,317.70 – 1,376.17

421 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: mistral-small-2506
Evaluation type: human preference

Run it locally

Inference · memory estimate

Mistral Small 3.2

Not enough verified architecture information for a reliable VRAM estimate.

API cost

USD / month

Input tokens / monthOutput tokens / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

Mistral Small 3.2

Provider offer

Capabilities

Context window: 128,000 tokens
Maximum output: 16,384 tokens
Reasoning: No
Tool calling: Yes
Open weights: Yes
Knowledge cutoff: 2025-03
Model license: Apache 2.0

Input modalities

text
image

Output modalities

text

* Provisional data