Home/Models/Step 3.7 Flash
stepfun logo

Step 3.7 Flash

by stepfun · Released May 2026

Open SourceMultimodal
Compare
Context
256K tokens (~128 books)
Input $/1M
$0.20
Output $/1M
$1.15
Type
multimodal
License
Open Source
Benchmarks
3 tested
Data updated today
About

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...

Tested on 3 benchmarks with 0.0% average. Top scores: Artificial Analysis — Coding Index (37.3%), Artificial Analysis — Quality Index (29.7%), Artificial Analysis — Agentic Index (21.5%).

Capabilities
speed
49.2
#55 globally
Benchmark Scores
Compare All
Tested on 3 benchmarks · Ranked across 1 categories
Score Distribution (all 274 models)
0255075100
Artificial Analysis — Coding Index

Artificial Analysis Coding Index. Composite coding quality score from multiple code benchmarks.

37.3
Artificial Analysis — Quality Index

Artificial Analysis Quality Index. Composite quality score combining multiple benchmark results into a single metric.

29.7
Artificial Analysis — Agentic Index

Artificial Analysis Agentic Index. Composite score measuring agent capability across tool use and planning tasks.

21.5
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
Community
BenchGecko API
step-3-7-flash
Specifications
  • Typemultimodal
  • Context256K tokens (~128 books)
  • ReleasedMay 2026
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.002
Available On
stepfun logostepfun$0.20
Categories
Share & Export
Tweet
Step 3.7 Flash is an open-source multimodal AI model by stepfun, released in May 2026. Context window: 256K tokens.