SWE-bench (full test set) leaderboard
- Claude Opus 4.5 leads
- 52.6%
- models measured
- 3
- lead over #2
- +33.8 pts
- median
- 18.8%
Leader on SWE-bench (full test set) · read October 21
# SWE-bench (full test set), read October 21
leaders = {
"Claude Opus 4.5": 52.6%, "GPT-4o": 18.8%, "GPT-4": 12.5%,
}
On October 21, 2026, 6 of our catalog's models had a SWE-bench (full test set) result, and Anthropic Claude Opus 4.5 led with 52.6%, ahead of Anthropic Claude 3.7 Sonnet at 33.8%.
The full SWE-bench test set of 2,294 GitHub issues from open source Python projects, before any curation. The number is the share of issues resolved.
Role in the index
A coding board we show beside the Vibe Coding Index, not inside it: the index keeps the six boards it was defined on, so a score does not move every time a board is added. Read the methodology
| # | Model | Result | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.5Anthropic | 52.6% | 44.0 | $25.00 | |
| 2 | Claude 3.7 SonnetRetiredAnthropic | 33.8% | Not benchmarked | $15.00 | |
| 3 | Claude 3.5 SonnetRetiredAnthropic | 24.9% | Not benchmarked | $15.00 | |
| 4 | GPT-4oOpenAI | 18.8% | Not benchmarked | $10.00 | |
| 5 | GPT-4OpenAI | 12.5% | Not benchmarked | $60.00 | |
| 6 | Claude 3 OpusRetiredAnthropic | 10.5% | Not benchmarked | $75.00 |
Only models in our catalog are listed, and each appears once with its best published run. 42 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.