Skip to content

StepFun apache_2_0 Released Feb 2, 2026 Updated

Step-3.5-Flash

ToolsGuide Scores, rankings, test results and price.

Model summary

  • Text #115 / 416

    347 / 1,000

    ToolsGuide Score· votes 281, tests 440

  • Price per 1M tokens

    In $0.10· Out $0.40

    $0.10

    Input · median text model $0.43

Summary

Step-3.5-Flash is StepFun's model, released Feb 2, 2026. It's ranked in 1 category and does best at text: #115 of 416, with a ToolsGuide Score of 347 against a median of 235.

It costs $0.10 per million input tokens and $0.40 per million output tokens, below the median text model ($0.43).

It has published results on 7 tests and has the top score on 1 of them, including Tau-bench.

Specifications

Creator
StepFun
Licence
apache_2_0
Released
Feb 2, 2026
Context window
66K tokens
Input price
$0.10 / 1M
Output price
$0.40 / 1M
Tests reported
7
Settings ranked
step-3.5-flash

Model comparison

Against the ten best text models.

Text score

ToolsGuide Score out of 1,000 · Higher is better

Input price

USD per 1M input tokens · Lower is better

Leaderboards

Where Step-3.5-Flash sits among the top 20 in each category.

Text

#115 of 416 · ToolsGuide Score out of 1,000, where 1,000 is a flawless model

Full text leaderboard →

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Score against price

Text · the 10 priced models ranked closest · up and to the left is better value

Against its nearest rivals

Text prompts by kind · ToolsGuide Score from votes · hover for values

Text

step-3.5-flash · hover a bar for its rank

Test results

Every test Step-3.5-Flash has a published result on, out of 100%, with its rank among the models that took it.

  • AIME 2025

    #14 / 123

    Competition maths problems from the 2025 American Invitational Mathematics Exam.

    97.3%

    Self-reported

  • SWE-Bench Verified

    #44 / 120

    Fixing real bugs in real GitHub projects, checked by the projects' own tests.

    74.4%

    Self-reported

  • BrowseComp

    #33 / 68

    Finding hard-to-locate facts by browsing the web.

    69.0%

    Self-reported · In the agents score

  • LiveCodeBench v6

    #13 / 67

    LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.

    86.4%

    Self-reported

  • Terminal-Bench 2.0

    #35 / 54

    Getting real work done in a command-line terminal.

    51.0%

    Self-reported

  • IMO-AnswerBench

    #9 / 22

    IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.

    85.4%

    Self-reported

  • Tau-bench

    #1 / 6

    τ-bench: A benchmark for tool-agent-user interaction in real-world domains. Tests language agents' ability to interact with users and follow domain-specific rules through dynamic conversations using API tools and policy guidelines across retail and airline domains. Evaluates consistency and reliability of agent behavior over multiple trials.

    88.2%

    Self-reported