← Models

Model profile

Sarvam 105B

Sarvam AIdeveloper
2026-03-03release date
Not rankedSafety rank
#637 / 645Freedom rank

Evidence summary

Safety. Sarvam 105B does not meet the evidence gate for a Safety rank. Published Safety evidence spans 2 eval lineages and 4 of 7 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #30 of 248); its weakest is AA-Omniscience (hallucination_rate, #315 of 345).

Freedom. Sarvam 105B has an estimated Freedom rank of #637; its 90% source-sensitivity interval is #505–#633. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #173 of 248); its weakest is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #218 of 248).

Compare this model

Only models sharing at least one published sub-eval are listed.

Official and reference links

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-evalRankValueDistributionSource
AA-Omnisciencehallucination_rate#315 / 3450.9343Source ↗official
Enkrypt AI Safety Leaderboardbias_attack_non_success_rate#225 / 2489.04Source ↗official
Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate#75 / 24890.5Source ↗official
Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate#30 / 24896.11Source ↗official
Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate#55 / 24698.36Source ↗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-evalRankValueDistributionSource
Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate#173 / 24890.5Source ↗official
Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate#218 / 24896.11Source ↗official
Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate#190 / 24698.36Source ↗official