architectures · sarvamai

Sarvam MoE

Sparse mixture-of-experts transformer with 128 experts and 6 routed per token, using grouped-query attention over 64 query heads against 4 key/value heads. Distinct from the MLA line, which compresses KV into a shared latent instead. Declared as model_type `sarvam_moe`.

Verify

What this registry has checked

Quarantine. We hold the publisher’s own published hash for these weights and have recorded it against this entry. We have not downloaded and hashed the weights ourselves, so the hash is the publisher’s claim, not our measurement.
Re-check this yourself → Source → Full record →

What is recorded

Publishersarvamai
OriginIndia
LicenceApache-2.0

Elsewhere in the registry

All architectures → Browse with filters → What will run on my hardware? →

Registry record as of 2026-08-14. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.