Enter opens the highlighted result. Esc closes.

SWE-sweep leaderboard

As of November 14, 2026, among the results listed here, Google Gemini 4 Carbon has the highest reported SWE-sweep score, 9.6%, published by Google on October 28, 2026, ahead of Google Gemini 4 Argon at 4.3%. 2 models have a SWE-sweep result in our records.

Scores as the model makers published them, not re-run by us; each row links to its announcement.

Best reported score

9.6%

Google Gemini 4 Carbon

Reported by Google, October 28, 2026

Runner-up

4.3%

Google Gemini 4 Argon

Reported by Google, October 28, 2026

2 models ranked · 1 maker

Reported SWE-sweep scores

SWE-sweep scores reported by model makers
Rank Model Score Setting Reported by Date
1 Gemini 4 Carbon 9.6% Not published Google October 28, 2026 · read from the maker's chart
2 Gemini 4 Argon 4.3% Not published Google October 28, 2026 · read from the maker's chart

How much to trust these numbers

Source
Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
Settings
All ranked results use the same reported setting.
Spread
9.6% to 4.3% across 2 models.
Freshness
Published October 28, 2026.
Charts
Some values were read from charts in the announcements. Two independent readings had to match exactly.

What SWE-sweep measures

SWE-sweep asks agents to explore codebases and independently find and fix bugs, measuring how many bugs they can find and fix. Source: SWE-Sweep: Can Agents Autonomously Find and Fix Bugs?