Step-SWE-Bench

Models fix software issues inside several coding agent harnesses, such as Claude Code, OpenCode and Hermes Agent. The score is the average over harnesses.

StepFun calls it in-house. The post gives per-harness scores for six harnesses and does not describe the tasks.

Made by StepFun Released May 2026 Fixing real repositories Tasks kept private Last in a launch post May 2026 , dropped before it saturated Last model added May 2026

Level: Expert Specialists with years of training, often over hours of work
Scored: Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers: Agent task Many steps in a terminal, browser, desktop, app, game or robot.

static.stepfun.com

Model scored in lab cards

Launch post lists it, from 1 lab

Step 3.7 Flash leads with 67.1%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

Bars Table

67.1 Step 3.7 Flash

1 Step 3.7 Flash 67.1%

#	Model	Score	Setting	Date	Source
1	Step 3.7 Flash	67.1%		May 2026	Model card

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Reported by the lab in its model card

Labs headline it less often now, last in May 2026

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

StepFun

Date	Launch post	Score	As the post writes it
29 May 2026	Step 3.7 Flash	67.1	Step-SWE-Bench avg over six harnesses

Step 3.7 Flash leads with 67.1%

Scores over time

Labs headline it less often now, last in May 2026

More in Coding