architectures · stepfun

Step Architecture (StepFun)

StepFun Step-series architecture. Sparse MoE with dense expert pre-training followed by MoE distillation. Step-3.5-Flash (196B) at 320 tok/s targets efficient math + science reasoning with 65.5K context. Step-3 (flagship) uses a fully MoE backbone with extended STEM harness benchmarking.

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.

Weight file SHA-256	da85a91891fe1c0f81853914019323a61d73e90b9d335d264c567741d9287fcf

Re-check this yourself → Source → Full record →

What is recorded

Publisher	stepfun
Origin	Recorded as a covered nation under 10 U.S.C. § 4872(f) — the PRC, Russia, Iran or North Korea. The registry did not record which.
Licence	StepFun-Open
Pulls recorded here	159,403

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.