Enterprise DNA
L Models Best vision

Step 3.7 Flash

by Stepfun

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

Visit Stepfun Submit your build →
S3

Models

Step 3.7 Flash

Added 10 July 2026

#stepfun

Overview

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...

How to use / API access

Call via OpenRouter with model id `stepfun/step-3.7-flash`.

  • openrouter · stepfun/step-3.7-flash Docs

OpenRouter id: stepfun/step-3.7-flash

Benchmarks

  • Intelligence Index (artificial-analysis) : 30.3 index
  • Coding Index (artificial-analysis) : 39.6 index
  • Agentic Index (artificial-analysis) : 21.5 index
  • Design Arena 3d (other) : 1195 elo · #48
  • Design Arena asciiart (other) : 1208 elo · #20
  • Design Arena codecategories (other) : 1211 elo · #46
  • Design Arena dataviz (other) : 1208 elo · #44

Notes

Step 3.7 Flash is listed in the EDNA Models directory from the OpenRouter catalogue.

← Back to Models Submit your own entry →