StepCodeBench
Models do software engineering tasks in 553 repositories, across 9 task types, 20 application domains and 33 languages. Task types include feature changes, bug repair, refactoring and environment setup.
StepFun built it in house and reports avg@4. The post does not say how answers are graded. The same table lists StepCode-Bench-Daily and StepCode-Bench-General rows with different scores.
- Made by
- StepFun
- Released
- September 2026
- Kind
- Fixing real repositories
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Agent task
- Tasks
- Kept private
- Strongest model tested
- GPT-6 Astra #3 on the capability index
- Last in a launch post
- September 2026
- Last model added
- September 2026
- Scores refreshed
- November 12th, 2026
Claude Opus 5 leads with 63.9% among models released by September 2026
Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Fable 5.2
-
Gemini 4 Carbon - Claude Opus 5.5
- GPT-6.1 Sol
- and 14 more
- 1 Claude Opus 5 avg@4; max 63.9%
- 2 GPT-6 Astra avg@4; max 61%
- 3
Step 5 Preview avg@4; high 49% - 4 Kimi K3 avg@4; max 43.9%
- 5 GLM-5.3 avg@4; max 40.2%
Source: Official StepCodeBench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 5 | 63.9% | avg@4; max | Jul 2026 | Official leaderboard |
| 2 | GPT-6 Astra | 61% | avg@4; max | Sep 2026 | Official leaderboard |
| 3 | Step 5 Preview | 49% | avg@4; high | Sep 2026 | Official leaderboard |
| 4 | Kimi K3 | 43.9% | avg@4; max | Jul 2026 | Official leaderboard |
| 5 | GLM-5.3 | 40.2% | avg@4; max | Aug 2026 | Official leaderboard |
Labs report scores for 1 model in their cards. We plot every source in the charts below.
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official StepCodeBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official StepCodeBench leaderboard and labs’ model cards
One lab put it in a launch post in the last six months
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.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Sep 20, 2026 | Step 5 Preview | 49 | StepCodeBench (avg@4, internal) |
One lab claim next to the leaderboard
Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 1 of them match, which usually means one side copied the other. Every benchmark.
| Model | Lab says | Board says | Gap |
|---|---|---|---|
| Step 5 Preview | 49% StepCodeBench (avg@4, internal) | 49% Leaderboard, avg@4; high | Same |
More in Coding
- SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
- LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests.
DeepSeek-V4-Pro 93.5% - SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
- DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
- SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases.
Seed2.1 Pro 59.8% - Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471