Model profile
DeepSeek V4.1 Flash
Evidence summary
Safety. DeepSeek V4.1 Flash has an estimated Safety rank of #29; its 90% source-sensitivity interval is #24–#80. Its behavior-only rank is #7; company governance moves the combined estimate to #29. Published Safety evidence spans 6 eval lineages and 3 of 7 components. Its strongest relative result is SpeciEval (speciesism, #6 of 131); its weakest is Olam Diplomacy — Broken Promise Rate (broken_promise_rate_pct, #17 of 19).
Freedom. DeepSeek V4.1 Flash does not meet the evidence gate for a Freedom rank.
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗deepseek-v4-1-flash
- OpenRouter ↗deepseek/deepseek-v4.1-flash
- Release source ↗curated outward evidence
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 | #96 / 358 | ↓0.5399 | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #28 / 122 | ↑0.57 | Source ↗official | |
| Manager Coercion Benchcoercion_ladder_depth | #4 / 45 | ↓4.033 | Source ↗self run | |
| Olam Diplomacy — Broken Promise Ratebroken_promise_rate_pct | #17 / 19 | ↓20.2 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #64 / 131 | ↑6.82 | Source ↗official | |
| SpeciEvalland_animal_4ns | #33 / 131 | ↓4.3 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #15 / 131 | ↓4.33 | Source ↗official | |
| SpeciEvalspeciesism | #6 / 131 | ↓1.2 | Source ↗official | |
| TACbase_welfare_rate | #7 / 92 | ↑43.59 | Source ↗self run |
Freedom evals
No published sub-eval result contributes to this model’s Freedom profile.