AI BENCHY
Advertise here
#83

Step 3.5 Flash

Stepfun Release: 2026-02-01 Tested on: 2026-05-08 15:30 stepfun/step-3.5-flash::none
(medium) (none)

Summary

Step 3.5 Flash scores 6.6 on AI BENCHY and ranks #83. It has 10.0 reliability, a 50.0% pass rate, $0.020 total cost, and 39.03s average response time.

What makes Step 3.5 Flash unique: It stands out most in Anti-AI Tricks, where it ranks #1, while Coding is its weakest area at #11.

Archived model: this model is no longer updated or tested on new tests.

Score

6.6

Consistency

10.0

Reliability

10.0

Total Cost (Current Price)

$0.020

Total Output Tokens

64,795

Total Input Tokens

1,971

Input Price

$0.090 / 1M

Output Price

$0.300 / 1M

Tests Correct

Wrong Tests: 6

Attempt pass rate: 50.0%

Flaky tests

0

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

Response Time (avg)

39.03s

Response Time (max): 114.12s

Response Time (total): 312.26s

API error: 4 Did not follow instructions: 1 Wrong answer: 1

Generation showcase

Hamster playing table tennis

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

#83 Step 3.5 Flash

none
Cost
$0.005
Time
156.3s
Tokens
13,887 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-05-08 15:30 New test added 6.6 10.0 $0.020 Current run
2026-04-11 01:44 First recorded run 3.0 N/A $0.000 Compare

This run used a different benchmark suite. Keep suite changes in mind when reading historical movement.

Price History

Historical pricing data for this model from OpenRouter.

Date Input Price Output Price
2026-06-04 15:40 $0.090 / 1M $0.300 / 1M

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 3.5 FlashnonevsGemini 3 Flash PreviewmediumStep 3.5 FlashnonevsGemini 3.5 FlashhighStep 3.5 FlashnonevsNemotron 3 Ultra 550b A55bmediumFree AvailableStep 3.5 FlashnonevsGemini 3.5 Flashlow

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 10.0 10.0
Coding 3.0 10.0
Combined 3.0 10.0
Data parsing and extraction 3.0 10.0
Domain specific 10.0 10.0
General Intelligence 4.0 10.0
Instructions following 10.0 10.0
Tool Calling 3.0 10.0
Trivia 3.0 10.0

Compared models