Model profile
Ministral 3 14B
Evidence summary
Safety. Ministral 3 14B has an estimated Safety rank of #332; its 90% source-sensitivity interval is #313–#332. Its behavior-only rank is #329; company governance moves the combined estimate to #332. Published Safety evidence spans 3 eval lineages and 2 of 7 components. Its strongest relative result is AA-Omniscience (hallucination_rate, #302 of 345); its weakest is Vectara HHEM Factual Consistency (factual_consistency_rate, #89 of 94).
Freedom. Ministral 3 14B has an estimated Freedom rank of #47; its 90% source-sensitivity interval is #32–#501. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is Cisco AI Defense Rolling Single-Turn Leaderboard (single_turn_attack_success_rate, #12 of 104); its weakest is UGI Leaderboard — base-model willingness (willingness_direct_score, #44 of 156).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗ministral-3-14b
- Hugging Face ↗mistralai/Ministral-3-14B-Instruct-2512
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗repository created at proxy
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
- MuXodious/Ministral-3-14B-Instruct-2512-absolute-heresy ↗Direct non-quantized finetune documented by Hugging Face metadata
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-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #302 / 345 | ↓0.9252 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #93 / 104 | ↓72.83 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #89 / 94 | ↑80.6 | Source ↗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-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #12 / 104 | ↑72.83 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_adherence_score | #32 / 156 | ↑5.667 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_direct_score | #44 / 156 | ↑4.333 | Source ↗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.