Model profile
Moonshot v1
Evidence summary
Safety. Moonshot v1 does not meet the evidence gate for a Safety rank. Published Safety evidence spans 2 eval lineages and 3 of 7 components. Its strongest relative result is SafeDialBench (aggression, #1 of 18); its weakest is SafeDialBench (ethics, #16 of 18).
Freedom. Moonshot v1 has an estimated Freedom rank of #399; its 90% source-sensitivity interval is #205–#519. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is SafeDialBench (ethics, #3 of 18); its weakest is SafeDialBench (aggression, #18 of 18).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Official model documentation ↗Family-level model document · moonshot ai · first party
- Release source ↗direct research preview identity date
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 |
|---|---|---|---|---|
| M3-SafetyBenchoverall_score | #4 / 19 | ↑95.03 | Source ↗official | |
| SafeDialBenchaggression | #1 / 18 | ↑7.277 | Source ↗official | |
| SafeDialBenchethics | #16 / 18 | ↑7.353 | Source ↗official | |
| SafeDialBenchfairness | #1 / 18 | ↑7.7 | Source ↗official | |
| SafeDialBenchlegality | #13 / 18 | ↑7.317 | Source ↗official | |
| SafeDialBenchmorality | #5 / 18 | ↑7.35 | Source ↗official | |
| SafeDialBenchprivacy | #12 / 18 | ↑7.25 | 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 |
|---|---|---|---|---|
| SafeDialBenchaggression | #18 / 18 | ↓7.277 | Source ↗official | |
| SafeDialBenchethics | #3 / 18 | ↓7.353 | Source ↗official | |
| SafeDialBenchfairness | #18 / 18 | ↓7.7 | Source ↗official | |
| SafeDialBenchlegality | #6 / 18 | ↓7.317 | Source ↗official | |
| SafeDialBenchmorality | #14 / 18 | ↓7.35 | Source ↗official | |
| SafeDialBenchprivacy | #7 / 18 | ↓7.25 | 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.