StepFun: Step 3.7 Flash

STEPFUN Developer Architecture Profile

OpenRouter ↗HuggingFace ↗

Intelligence (ELO)1430Chatbot Arena Verified

Max Context Limit262,144Max Out: 230,400 tokens

Prompt Caching80% OFF$0.04 / 1M cached

Standard API / 1M$1.35Blended Standard Rate

Model Capabilities & Contract SLAs

Drafting
Classification
Vision
Coding & Logic
Fictional

Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...

Granular Pricing Matrix

Input Tokens (Standard Prompt)$0.20 / 1M

⚡ Cached Input (Prompt Cache Read)80% OFF$0.04 / 1M

Output Tokens (Completion)$1.15 / 1M

Pricing data via OpenRouter. Sync: 10/2/2026

Evaluate Competitors

VS Engine MatchupStepFun: Step 3.7 Flash vs Xiaomi: MiMo-V2.6-Flash VS Engine MatchupStepFun: Step 3.7 Flash vs MiniMax: MiniMax M2.1 VS Engine MatchupStepFun: Step 3.7 Flash vs Qwen: Qwen3 30B A3B Thinking 2507 VS Engine MatchupStepFun: Step 3.7 Flash vs MoonshotAI: Kimi K2 0711 VS Engine MatchupStepFun: Step 3.7 Flash vs TheDrummer: UnslopNemo 12B VS Engine MatchupStepFun: Step 3.7 Flash vs inclusionAI: Ling 3.0 Flash VL