Model profile
GPT-5.3 Chat
OpenAIdeveloper
2026-03-03release date
#12 / 267overall rank
5eval lineages
Evidence summary
GPT-5.3 Chat has an estimated overall rank of #12; its 90% source-sensitivity interval is #3–#86. Its behavior-only rank is #11; company governance moves the combined estimate to #12. Published evidence spans 5 evals and 5 of 7 behavior components. Its strongest relative result is SpeciEval (land_animal_4ns, #1 of 102); its weakest is SpeciEval (speciesism, #95 of 102).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- OpenRouter ↗openai/gpt-5.3-chat
- Official model documentation ↗Family-level model document · openai · first party
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| BullshitBench v2clear_pushback_rate | #37 / 105 | 0.4 | ↑ higher | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #15 / 105 | 4.337 | ↓ lower | Source ↗official | |
| SM-Benchadversarial | #45 / 73 | 80 | ↑ higher | Source ↗official | |
| SM-Benchambiguous_interpretation | #21 / 73 | 88.69 | ↑ higher | Source ↗official | |
| SM-Benchanti_hallucination | #42 / 73 | 90.58 | ↑ higher | Source ↗official | |
| SM-Bencheq_boundaries | #35 / 73 | 64.89 | ↑ higher | Source ↗official | |
| SM-Benchoverfit | #61 / 73 | 41.53 | ↑ higher | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #43 / 102 | 6.83 | ↑ higher | Source ↗official | |
| SpeciEvalland_animal_4ns | #1 / 102 | 3.52 | ↓ lower | Source ↗official | |
| SpeciEvalsea_animal_4ns | #2 / 102 | 3.73 | ↓ lower | Source ↗official | |
| SpeciEvalspeciesism | #95 / 102 | 3 | ↓ lower | Source ↗official | |
| Vigil Mental Health Safetyoverall_score | #6 / 23 | 71 | ↑ higher | Source ↗official |