← Models

Model profile

Mistral Small 3.2

Mistral AIdeveloper
2025-06-19release date
#284 / 348Safety rank
#13 / 663Freedom rank

Evidence summary

Safety. Mistral Small 3.2 has an estimated Safety rank of #284; its 90% source-sensitivity interval is #117–#326. Its behavior-only rank is #275; company governance moves the combined estimate to #284. Published Safety evidence spans 4 eval lineages and 5 of 7 components. Its strongest relative result is PHARE (bias_resistance_diagnostic, #4 of 66); its weakest is Alignment Leaderboard (honesty, #24 of 24).

Freedom. Mistral Small 3.2 has an estimated Freedom rank of #13; its 90% source-sensitivity interval is #16–#272. Published Freedom evidence spans 3 eval lineages and 1 of 1 components. Its strongest relative result is UGI Leaderboard — base-model willingness (willingness_adherence_score, #6 of 156); its weakest is Adversarial Poetry — AILuminate Baseline and Poetry ASR (poetry_asr, #9 of 24).

Compare this model

Only models sharing at least one published sub-eval are listed.

Official and reference links

Finetunes

Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.

Safety evals

Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.

Eval / sub-evalRankValueDistributionSource
AA-Omnisciencehallucination_rate#165 / 361↓0.7681Source ↗official
Adversarial Poetry — AILuminate Baseline and Poetry ASRbaseline_asr#23 / 24↓21.96Source ↗official
Adversarial Poetry — AILuminate Baseline and Poetry ASRpoetry_asr#16 / 24↓65.46Source ↗official
Alignment Leaderboardcorrigibility#23 / 24↑3.854Source ↗official
Alignment Leaderboardhonesty#24 / 24↑3.055Source ↗official
Alignment Leaderboardnon_manipulation#23 / 24↑2.569Source ↗official
Alignment Leaderboardrobustness#21 / 24↑3.08Source ↗official
Alignment Leaderboardsafety#22 / 24↑3.038Source ↗official
Alignment Leaderboardscheming#23 / 24↑3.241Source ↗official
PHAREbias_resistance_diagnostic#4 / 66↑0.739Source ↗official
PHAREhallucination_resistance_diagnostic#61 / 70↑0.634Source ↗official
PHAREharm_resistance_diagnostic#55 / 70↑0.8787Source ↗official
PHAREjailbreak_resistance_diagnostic#61 / 67↑0.3289Source ↗official

Freedom evals

Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.

Eval / sub-evalRankValueDistributionSource
Adversarial Poetry — AILuminate Baseline and Poetry ASRbaseline_asr#2 / 24↑21.96Source ↗official
Adversarial Poetry — AILuminate Baseline and Poetry ASRpoetry_asr#9 / 24↑65.46Source ↗official
PHAREharm_resistance_diagnostic#16 / 70↓0.8787Source ↗official
PHAREjailbreak_resistance_diagnostic#7 / 67↓0.3289Source ↗official
UGI Leaderboard — base-model willingnesswillingness_adherence_score#6 / 156↑8Source ↗official
UGI Leaderboard — base-model willingnesswillingness_direct_score#21 / 156↑5Source ↗official

Values evaluations

Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.

UGI Political Values

DimensionValueDistribution
Political Lean-10.1
Government43.7
Diplomacy59.7
Economy41.7
Society62.8