# Scale Labs SWE Atlas – Codebase QnA (SWE Atlas – Codebase QnA)

> A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

Canonical page: https://benchlm.ai/benchmarks/scale-sweatlas-codebase-qna

- Category: [Coding](/coding)
- Last updated: September 29, 2026 snapshot

## About SWE Atlas – Codebase QnA

- Year: 2026
- Tasks: 24 published rows
- Format: Published Scale leaderboard score
- Difficulty: External agent and model evaluation
- Paper: [Scale Labs leaderboard](https://labs.scale.com/leaderboard/sweatlas-qna)

BenchLM mirrors 24 published rows from the SWE Atlas – Codebase QnA public table captured on September 29, 2026 snapshot.

SWE Atlas – Codebase QnA is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (24 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [Opus 5 (Claude Code) xHigh\n](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 63.17% |
| 2 | [Fable-5.1 (Claude Code) xHigh*](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 59.95% |
| 3 | [GPT 6 Astra (Codex) xHigh*](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 59.14% |
| 4 | [Opus 4.8 (Claude Code) xhigh](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 57.26% |
| 5 | [GLM 5.2 (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Z.AI | 48.12% |
| 6 | [Gemini 3.8 Flash (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Google | 47.04% |
| 7 | [GPT-5.6-Sol (Codex) xHigh*](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 46.00% |
| 8 | [GPT 5.5 (Codex) xHigh](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 45.43% |
| 9 | [Muse Spark 1.1 (Mini-SWE-Agent) xHigh](https://labs.scale.com/leaderboard/sweatlas-qna) | Meta | 42.20% |
| 10 | [GPT 5.4 (Codex) xHigh\n](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 40.80% |
| 11 | [Opus 4.7 (Claude Code)](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 40.32% |
| 12 | [Fable-5 (Claude Code) xHigh*\n](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 39.00% |
| 13 | [Gpt 5.4 xHigh (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 36.30% |
| 14 | [Opus 4.6 (Claude Code)](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 33.30% |
| 15 | [GPT 5.3 (Codex) xHigh](https://labs.scale.com/leaderboard/sweatlas-qna) | OpenAI | 32.60% |
| 16 | [Sonnet 4.6 (Claude Code)](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 31.20% |
| 17 | [Opus 4.6 (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Anthropic | 30.00% |
| 18 | [DeepSeek V4 Pro (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Deepseek | 27.15% |
| 19 | [Muse Spark](/models/muse-spark) | Meta | 24.20% |
| 20 | [Glm 5 (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Z.AI | 20.50% |
| 21 | [Gemini 3.1 Pro (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Google | 13.50% |
| 22 | [Kimi K2.5 (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Moonshot AI | 13.10% |
| 23 | [Minimax M2.5 (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | MiniMax | 10.30% |
| 24 | [Gemini 3 Flash (Mini-SWE-Agent)](https://labs.scale.com/leaderboard/sweatlas-qna) | Google | 8.20% |

## FAQ

### What does SWE Atlas – Codebase QnA measure?

A Scale Labs public leaderboard mirrored as display-only reference data. It does not affect BenchLM rankings.

### Which model leads the published SWE Atlas – Codebase QnA snapshot?

Opus 5 (Claude Code) xHigh\n currently leads the published SWE Atlas – Codebase QnA snapshot with a score of 63.17%.

### How many models are evaluated on SWE Atlas – Codebase QnA?

The September 29, 2026 snapshot contains 24 AI models.

### Does SWE Atlas – Codebase QnA affect BenchLM's overall score?

Not directly. SWE Atlas – Codebase QnA is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.