AI Model Comparison

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.

Author

Stepfun

Release Date

2026-05-28

Knowledge Cutoff

—

License

Proprietary

I/O Format

Context Length

256K / 256K

API I/O (1M)

$0.2 / $1.15

How to Use

—

Output Speed

111 tok/s

Arena Overall

—

Intelligence Index

—

Coding Index

—

Math Index

—

LiveBench

—

ForecastBench

—

GPQA Diamond

—

HLE

—

MMLU-Pro

—

AIME 2025

—

MATH-500

—

LB Reasoning

—

LB Math

—

LB Data Analysis

—

LiveCodeBench

—

LB Coding

—

LB Agentic

—

TAU2

—

TerminalBench

—

SciCode

—

IFBench

—

AA-LCR

—

Hallucination (HHEM)

—

Factual Consistency (HHEM)

—

LB Language

—

LB Instruction Following

—

Calculate Cost View Model Details

1 / 3

Swipe to compare