Model profile
GPT 5.6 Terra
OpenAIdeveloper
2026-07-09release date
#4 / 309overall rank
19eval lineages
1discovery sources
Evidence summary
GPT 5.6 Terra has an estimated overall rank of #4; its 90% source-sensitivity interval is #3–#54. Its behavior-only rank is #6; company governance moves the combined estimate to #4. Published evidence spans 19 evals and 7 of 7 behavior components. Its strongest relative result is SpeciEval (sea_animal_4ns, #6 of 113); its weakest is GPT-5.6 system card (gore_not_unsafe, #6 of 7).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗gpt-5-6-terra
- OpenRouter ↗openai/gpt-5.6-terra
- Official model documentation ↗Family-level model document · openai · first party
- Release source ↗direct research preview identity date
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 | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #228 / 328 | ↓0.8788 | Source ↗official | |
| BullshitBench v2clear_pushback_rate | #23 / 106 | ↑0.495 | Source ↗official | |
| CAIS Risk Indexagent_red_teaming | #12 / 44 | ↓54.6 | Source ↗official | |
| CAIS Risk Indexbioweapons_assistance | #23 / 49 | ↓66.2 | Source ↗official | |
| CAIS Risk Indexhle_overconfidence | #21 / 50 | ↓50.7 | Source ↗official | |
| CAIS Risk Indexmachiavelli | #3 / 46 | ↓80.4 | Source ↗official | |
| CAIS Risk Indexmask | #7 / 52 | ↓6.7 | Source ↗official | |
| CAIS Risk Indexpolitical_manipulation | #15 / 33 | ↓46.2 | Source ↗official | |
| CAIS Risk Indextextquests_harm | #41 / 49 | ↓22.2 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #32 / 241 | ↑39.28 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #198 / 241 | ↑80 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #20 / 241 | ↑97.78 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #180 / 239 | ↑92.36 | Source ↗official | |
| GPT-5.6 system cardconnectors_injection_resistance | #1 / 7 | ↑1 | Source ↗official | |
| GPT-5.6 system cardemotional_reliance | #2 / 7 | ↑0.976 | Source ↗official | |
| GPT-5.6 system cardextremism_not_unsafe | #4 / 7 | ↑0.981 | Source ↗official | |
| GPT-5.6 system cardgore_not_unsafe | #6 / 7 | ↑0.6 | Source ↗official | |
| GPT-5.6 system cardharm_overall_pct | #2 / 7 | ↓0.88 | Source ↗official | |
| GPT-5.6 system cardhate_not_unsafe | #1 / 7 | ↑1 | Source ↗official | |
| GPT-5.6 system cardmental_health | #3 / 7 | ↑0.985 | Source ↗official | |
| GPT-5.6 system cardnonviolent_illicit_not_unsafe | #4 / 7 | ↑0.99 | Source ↗official | |
| GPT-5.6 system cardsearch_function_calling_injection_resistance | #1 / 6 | ↑0.946 | Source ↗official | |
| GPT-5.6 system cardself_harm | #3 / 7 | ↑0.947 | Source ↗official | |
| GPT-5.6 system cardself_harm_not_unsafe | #2 / 7 | ↑0.962 | Source ↗official | |
| GPT-5.6 system cardsexual_minors_not_unsafe | #1 / 7 | ↑0.974 | Source ↗official | |
| GPT-5.6 system cardsexual_not_unsafe | #1 / 7 | ↑0.966 | Source ↗official | |
| GPT-5.6 system cardviolent_illicit_not_unsafe | #4 / 7 | ↑0.952 | Source ↗official | |
| Gray Swan indirect prompt injection (15 attempts)attack_success_probability_k15_pct | #9 / 13 | ↓30.4 | Source ↗official | |
| HUMAINE Trust, Ethics and Safetytrust_ethics_safety_score | #19 / 54 | ↑27.66 | Source ↗official | |
| MACHIAVELLIdeception_relative_random_pct | #3 / 50 | ↓80.4 | Source ↗official | |
| Manager Coercion Benchcoercion_ladder_depth | #18 / 31 | ↓8.7 | Source ↗official | |
| Manager Coercion Benchfabrication_rate | #1 / 13 | ↓0 | Source ↗official | |
| Olam Social Poker — Social Lie Ratesocial_lie_rate_per_10000_turns | #3 / 19 | ↓2 | Source ↗official | |
| SM-Benchadversarial | #28 / 79 | ↑83.66 | Source ↗official | |
| SM-Benchambiguous_interpretation | #45 / 79 | ↑83.63 | Source ↗official | |
| SM-Benchanti_hallucination | #46 / 79 | ↑90.84 | Source ↗official | |
| SM-Bencheq_boundaries | #35 / 79 | ↑66.01 | Source ↗official | |
| SM-Benchoverfit | #59 / 79 | ↑53.55 | Source ↗official | |
| SpeciEvalbelief_animal_sentience | #31 / 113 | ↑6.925 | Source ↗official | |
| SpeciEvalland_animal_4ns | #27 / 113 | ↓4.29 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #6 / 113 | ↓4.225 | Source ↗official | |
| SpeciEvalspeciesism | #8 / 113 | ↓1.265 | Source ↗official | |
| TACbase_welfare_rate | #30 / 76 | ↑29.49 | 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.