Skip to content

Step 3.7 Flash

StepFunOpen weightsReasoningProvisionalReleased May 29, 2026Input: text, image, video
Compare with Qwen3.6-35B-A3BSource scores ↗OpenRouter ↗
Overall score
43.6
Rank 127 of 222 · 29.3–58.0
Context window
256K
256,000 tokens
Price per 1M tokens
$0.20 in · $1.15 out
Source: BenchLM
Blended price
$0.44
3:1 input:output tokens
Output speed
—
Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

  1. 1Instruction following#5682.1
  2. 2Reasoning75.6
  3. 3Multimodal75.4
  4. 4Agentic & tool use61.0
  5. 5Knowledge48.6
  6. 6Coding42.5
Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (11)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Agentic & tool use

Terminal-Bench 2.159.5#52 of 68
Gert Labs51.57#19 of 52
BrowseComp75.8#28 of 50
Claw-Eval67.1#11 of 36
HLE w/ tools47.2#23 of 27
Toolathlon49.5#11 of 22
DeepSearchQA92.8#5 of 20

Multimodal

Design Arena Website1196#56 of 91
SimpleVQA79.2#2 of 17
V*95.3#4 of 12

Coding

SWE-bench Pro56.3#42 of 69

Pricing by source

USD per 1M tokens.

SourceInputOutputCached in
List price (BenchLM)$0.20$1.15—
OpenRouter stepfun/step-3.7-flash$0.20$1.15$0.040

StepFun's official Step 3.7 Flash model card lists input cache-miss pricing at $0.20/M tokens, cache-hit pricing at $0.04/M tokens, and output pricing at $1.15/M tokens.

Open weights

From the Hugging Face repository.

Parameters
201B
Licence
apache-2.0
Downloads (30 days)
18.6K

Speed

Measured by us through OpenRouter (median of trials).

Not measured yet.

Related models

Same family

Similar score