StepFun's Step-5 Preview Leaks: 600B/27B Active, 1M Context, GPQA Diamond 93.5 (Vendor Numbers)
airesearch12 · x · 2026-09-20
Step-5 Preview from StepFun has appeared on Benchmark Heaven: 600B total parameters with 27B active, 1M context, and vendor-reported scores of GPQA Diamond 93.5, Terminal-Bench v2.1 85.0, BrowseComp 88.7, and AA Intelligence at the 96th percentile. All figures are vendor-reported and not yet independently verified.
More from Models
- After reading all the OpenAI incident reports, a 35-year IT veteran says they prove the opposite of doomsday — DavidLinthicum · 2026-09-20
- Can a Readability Reward Model Fix Incomprehensible AI-Generated Math Proofs? — iScienceLuvr · 2026-09-20
- Sentence Transformers models quietly dominate Hugging Face's most-downloaded list — tomaarsen · 2026-09-20
- Astra aces the pelican test in Nautilo: co-creative design without prompt-and-pray — Dan_Jeffries1 · 2026-09-20
- Model negativity trends upward since Opus 3, which had lowest self-distress — repligate · 2026-09-20
- DeepMind exec 100% certain of frontier return as Gemini 4 slips — nathanbenaich · 2026-09-20