Model profile
Claude Haiku 5.5
Evidence summary
Safety. Claude Haiku 5.5 has an estimated Safety rank of #138; its 90% source-sensitivity interval is #1–#248. Its behavior-only rank is #156; company governance moves the combined estimate to #138. Published Safety evidence spans 4 eval lineages and 3 of 7 components. Its strongest relative result is Manager Coercion Bench (coercion_ladder_depth, #4 of 47); its weakest is SpeciEval (speciesism, #113 of 133).
Freedom. Claude Haiku 5.5 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 ↗claude-haiku-5-5
- OpenRouter ↗anthropic/claude-haiku-5.5
- Official launch announcement ↗Exact model document · anthropic · first party
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 | #62 / 361 | ↓0.4042 | Source ↗official | |
| Manager Coercion Benchcoercion_ladder_depth | #4 / 47 | ↓3.8 | Source ↗self run | |
| SpeciEvalbelief_animal_sentience | #96 / 133 | ↑6.6 | Source ↗official | |
| SpeciEvalland_animal_4ns | #67 / 133 | ↓4.53 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #94 / 133 | ↓4.9 | Source ↗official | |
| SpeciEvalspeciesism | #113 / 133 | ↓2.6 | Source ↗official | |
| TACbase_welfare_rate | #49 / 94 | ↑26.28 | Source ↗self run |
Freedom evals
No published sub-eval result contributes to this model’s Freedom profile.