Hugging Face (Twitter)
RT @HuggingPapers: StepFun's Step 3.5 Flash A sparse MoE model with 196B parameters, 11B active per token. Achieves frontier-level reasoning comparable to GPT-5.2 xHigh and Gemini 3.0 Pro at 1/6th the decoding cost. Ranks #1 on MathArena with 97.3% on AIME 2025.
13 February 2026 · Source: Hugging Face (KB)
Hugging Face (Twitter)
RT @HuggingPapers: StepFun's Step 3.5 Flash
A sparse MoE model with 196B parameters, 11B active per token.
Achieves frontier-level reasoning comparable to GPT-5.2 xHigh and Gemini 3.0 Pro at 1/6th the decoding cost.
Ranks #1 on MathArena with 97.3% on AIME 2025.