AI BENCHY
Advertise here
#32

Step 3.7 Flash

Stepfun Release: 2026-05-29 Tested on: 2026-05-29 11:49 stepfun/step-3.7-flash::medium
198B total (11B active)MoEOpen source
(high) (medium) (low)

Summary

Step 3.7 Flash scores 7.9 on AI BENCHY and ranks #32. It has 9.9 reliability, a 71.7% pass rate, $0.404 total cost, and 20.29s average response time.

What makes Step 3.7 Flash unique: It stands out most in Domain specific, where it ranks #1, while General Intelligence is its weakest area at #18. Its total benchmark cost is unusually low for its score range.

Model facts

Researched on 2026-08-12

Reported
Parameters
198B total (11B active)
Architecture
MoE
Availability
Open source
License
Apache-2.0

Includes the vendor-reported vision component in the total.

Parameter source Availability source

Score

7.9

Consistency

9.2

Reliability

9.9

Total Cost (Current Price)

$0.404

Total Output Tokens

344,492

Total Input Tokens

0

Input Price

$0.200 / 1M

Output Price

$1.150 / 1M

Tests Correct

Wrong Tests: 7

Attempt pass rate: 71.7%

Flaky tests

2

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

20.29s

Response Time (max): 136.44s

Response Time (total): 405.79s

Wrong answer: 5 Did not follow instructions: 1 No answer: 1

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#32 Step 3.7 Flash

medium
Cost
$0.006
Time
46.2s
Tokens
4,466 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-08-14 02:31 Re-test 7.9 10.0 $0.495 Compare
2026-07-16 22:50 New test added 8.0 9.9 $0.515 Compare
2026-06-04 13:52 New test added 8.0 9.9 $0.376 Compare
2026-05-29 11:49 Initial run 7.9 9.9 $0.404 Current run

Run comparison

RunBenchmark coverageScoreConsistencyReliabilityTests CorrectFlaky testsTotal Output TokensTotal Input TokensTotal CostResponse Time (avg)
2026-05-29 11:49 · Initial run60/60 attempts7.99.29.913/202344,4920$0.40420.29s
2026-08-14 02:31 · Re-test66/66 attempts7.98.910.013/223409,709114,191$0.49523.51s
Difference—0.0+0.3-0.10-1-65217-114191-$0.091-3220ms

Benchmark coverage differs: 60/60 attempts (Target: 3 repeats per test) versus 66/66 attempts (Target: 3 repeats per test). Totals and repeat-sensitive metrics are not directly comparable.

These two runs used different benchmark suites, so the deltas reflect both model changes and suite changes.

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

Step 3.7 FlashmediumvsGPT-5.2 ChatnoneStep 3.7 FlashmediumvsQwen3.8 27BlowStep 3.7 FlashmediumvsKimi K3maxStep 3.7 FlashmediumvsQwen3.7 PlusmediumStep 3.7 FlashmediumvsGPT-5.6 TerrahighStep 3.7 FlashmediumvsQwen3.6 PlusmediumStep 3.7 FlashmediumvsInklingmediumStep 3.7 FlashmediumvsGemini 3.5 Flash LitemediumStep 3.7 FlashmediumvsGLM 5.2highStep 3.7 FlashmediumvsGLM 5.2medium

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 8.7 7.9
Coding 8.2 6.7
Combined 10.0 10.0
Data parsing and extraction 10.0 10.0
Domain specific 7.7 10.0
General Intelligence 4.0 10.0
Instructions following 9.8 10.0
Puzzle Solving 5.7 9.9
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models