SWE-Atlas Refactoring
We show this table for reference; we do not rank on it.
A Scale SWE-Atlas software-engineering agent benchmark focused on refactoring tasks.
Scale score on SWE-Atlas Refactoring — October 28, 2026 snapshot
We mirror the published scale score view for SWE-Atlas Refactoring. GPT 6 Astra (Codex) xHigh leads the public snapshot at 59.0%, followed by Fable-5.1 (Claude Code) xHigh (56.7%) and Fable-5 (Claude Code) xHigh (54.8%). We do not use these results to rank models overall.
GPT 6 Astra (Codex) xHigh
OpenAI
Fable-5.1 (Claude Code) xHigh
Anthropic
Fable-5 (Claude Code) xHigh
Anthropic
17 modelsAgenticCurrentDisplay onlyUpdated October 28, 2026 snapshot
Scale score table (17 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | GPT 6 Astra (Codex) xHighOpenAI | 59.0% | Not reported | Unknown |
| 2 | Fable-5.1 (Claude Code) xHighAnthropic | 56.7% | Not reported | Unknown |
| 3 | Fable-5 (Claude Code) xHighAnthropic | 54.8% | Not reported | Unknown |
| 4 | Opus-4.7 (Claude Code)Anthropic | 48.6% | Not reported | Unknown |
| 5 | Opus 4.8 (Claude Code)\nAnthropic | 46.7% | Not reported | Unknown |
| 6 | GPT-5.5 (Codex) xHighOpenAI | 44.8% | Not reported | Unknown |
| 7 | Gemini 3.8 Flash (Mini-SWE-Agent)Google | 44.8% | Not reported | Unknown |
| 8 | GPT 5.4 (Codex) xHighOpenAI | 44.3% | Not reported | Unknown |
| 9 | GLM 5.2 (Mini-SWE-Agent)Z.AI | 42.4% | Not reported | Unknown |
| 10 | GPT 5.3 (Codex) xHighOpenAI | 42.4% | Not reported | Unknown |
| 11 | Opus-4.6 (Claude Code)Anthropic | 35.6% | Not reported | Unknown |
| 12 | Gemini-3.1-Pro (Gemini CLI)Google | 33.8% | Not reported | Unknown |
| 13 | Sonnet-4.6 (Claude Code)Anthropic | 32.2% | Not reported | Unknown |
| 14 | Glm-5 (Mini-SWE-Agent)Z.AI | 24.2% | Not reported | Unknown |
| 15 | Kimi-K2.5 (Mini-SWE-Agent)Moonshot AI | 20.9% | Not reported | Unknown |
| 16 | Minimax-M2.5 (Mini-SWE-Agent)MiniMax | 19.5% | Not reported | Unknown |
| 17 | Gemini-3-Flash (Mini-SWE-Agent)Google | 10.0% | Not reported | Unknown |
How BenchLM shows SWE Atlas – Refactoring
BenchLM mirrors 17 published rows from Scale Labs’ public SWE Atlas – Refactoring leaderboard, captured on October 28, 2026 snapshot.
The table is display only. It is useful context for a published agent or model configuration, but it does not enter BenchLM’s overall or category rankings.
The published SWE-Atlas Refactoring snapshot places GPT 6 Astra (Codex) xHigh first at 59.0%. The third row is 4.3 points behind. The broader top-10 range is 16.7 points, so the table still separates the published systems.
17 models have been evaluated on SWE-Atlas Refactoring. The benchmark falls in the Agentic category. SWE-Atlas Refactoring is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About SWE-Atlas Refactoring
Year
2026
Tasks
SWE-Atlas refactoring tasks
Format
Refactoring score with confidence intervals
Difficulty
Real-world software-engineering agent tasks
BenchLM mirrors the public Scale SWE-Atlas Refactoring leaderboard as a display-only agentic software-engineering benchmark. The source compares model-agent combinations such as Claude Code, Codex, Gemini CLI, and Mini-SWE-Agent.
Freshness and provenance
Version
SWE-Atlas Refactoring 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does SWE-Atlas Refactoring measure?
A Scale SWE-Atlas software-engineering agent benchmark focused on refactoring tasks.
Which model leads the published SWE-Atlas Refactoring snapshot?
GPT 6 Astra (Codex) xHigh currently leads the published SWE-Atlas Refactoring snapshot with 59.0% scale score. BenchLM shows this benchmark for display only and does not use it in overall rankings.
How many models are evaluated on SWE-Atlas Refactoring?
The October 28, 2026 snapshot snapshot contains 17 AI models.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.