SWE-sweep
Subtask-weighted reward on long-term software modification tasks.
- Category
- Agentic coding
- Unit
- Percent (%)
- Higher is better
- Comparison groups
- 1total
- All model rankings (records) · Ranked results: 1
All model rankings (records)
These rankings are recalculated by comparison group including legacy models and models unavailable in Claude Code / Codex. They differ from the rankings above because the set of models differs.
Show rankings including legacy and unavailable models (entries: 1)
official results (Gemini 4 Carbon (Google DeepMind models page, benchmark table))
| Rank | Model | effort | Score | Source cost (metric) | measured/reported by | source |
|---|---|---|---|---|---|---|
| 1 | Gemini 4 Carbon | High | 9.6% | — | Gemini 4 Carbon (Google DeepMind models page, benchmark table)Retrieved 2026-11-12 · Vendor-reported value. Google published table. methodology PDF: Gemini 4 Carbon/Argon used the highest Gemini API thinking setting, pass@1, and a single trial. Row-specific source: self-measured for Carbon using the mini-swe-agent harnessSource excerptSWE-sweep: Gemini 4 Carbon 9.6 |