SEAL - SWE Atlas - Codebase QnA: leaderboard
Scale SEAL evaluation of LLM ability to answer questions about large codebases, testing code comprehension and navigation.
Metric: Score. Source: scale.com. Source: benchmarklist.com. Saturation forecast: Around July 2027. 24 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.8 Max (0902) | 66.3 | #29 |
| 2 | Kimi K3 | 66 | #30 |
| 3 | Claude Opus 5 (Max) | 66 | #12 (Claude Opus 5) |
| 4 | DeepSeek V4 Pro (0813) | 66 | #53 |
| 5 | Hy4 preview | 64 | #38 |
| 6 | Claude Opus 5 (xHigh) | 63.17 | #12 (Claude Opus 5) |
| 7 | Claude Opus 5 | 62 | #12 |
| 8 | Qwen 3.8 Max | 62 | #35 |
| 9 | GLM-5.3 Flash | 61.29 | #65 |
| 10 | GPT-6 Astra (Max) | 60.9 | #6 (GPT-6 Astra) |
| 11 | Claude Fable 5.1 (xHigh) | 59.95 | #9 (Claude Fable 5.1) |
| 12 | GLM-5.3 (Max) | 59.6 | #39 (GLM-5.3) |
| 13 | Kimi K3 (Max) | 59.5 | #30 (Kimi K3) |
| 14 | Muse Spark 1.3 (Max) | 59.4 | #24 (Muse Spark 1.3) |
| 15 | Mistral Large 4 | 59.4 | #94 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=seal-swe-atlas-codebase-qna · How It Works · Data refreshed daily, snapshot 2026-10-20.