SWE Atlas v1.1: Test Writing leaderboard
As of 16 Nov 2026, the top score on SWE Atlas v1.1: Test Writing is 78.9% by Claude Fable 5.1 (Anthropic), ahead of Claude Fable 5.2 at 77.8%, across 28 model configurations with a published result (Scale AI SEAL, result dated 6 Nov 2026). Writing tests for existing code, SWE Atlas v1.1. Scale AI. SWE Atlas v1.1: Test Writing is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 16 Nov 2026)
- Claude Fable 5.1 holds the top SWE Atlas v1.1: Test Writing score at 78.9% (Scale AI SEAL, 6 Nov 2026).
- Claude Fable 5.2 is second on SWE Atlas v1.1: Test Writing at 77.8% (Scale AI SEAL, 11 Nov 2026).
- SWE Atlas v1.1: Test Writing: Writing tests for existing code, SWE Atlas v1.1. Scale AI (Scale AI SEAL).
- 28 model configurations have a published SWE Atlas v1.1: Test Writing result on BenchLeader (BenchLeader, 16 Nov 2026).
- SWE Atlas v1.1: Test Writing does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Writing tests for existing code, SWE Atlas v1.1. Scale AI.
- Published by
- Scale AI SEAL
- Category
- Coding
- Index weight
- Reference only
- Models
- 28
- Data as of
- 16 Nov 2026
- Newest result
- 11 Nov 2026
Scale AI SEAL Leaderboards.
Not comparable with the earlier test of the same name, whose final scores stay on SWE Atlas v1.0: Test Writing.
- 1Claude Fable 5.178.9%
- 2Claude Fable 5.277.8%
- 3Gemini 4 Argon77.8%
- 4Claude Opus 5.577.8%
- 5Gemini 4 Carbon76.7%
- 6Gemini 3.8 Flash71.1%
- 7Claude Sonnet 5.567.8%
- 8GPT-6 Astra66.7%
- 9GPT-6.1 Sol66.7%
- 10Grok 4.764.4%
- 11Kimi K363.3%
- 12GPT-6.1 Terra63.3%
- 13Claude Haiku 5.563.3%
- 14Muse Wonder61.1%
- 15DeepSeek V4.1 Flash61.1%
| # | ||||
|---|---|---|---|---|
| 1 | 78.9% | 66.4 | 2026-11-06 | |
| 2 | New | 77.8% | 65.4 | 2026-11-11 |
| 3 | 77.8% | 64.4 | 2026-11-06 | |
| 4 | 77.8% | 61.0 | 2026-11-06 | |
| 5 | 76.7% | – | 2026-11-11 | |
| 6 | 71.1% | 56.4 | 2026-11-06 | |
| 7 | 67.8% | – | 2026-11-06 | |
| 8 | 66.7% | 66.4 | 2026-11-06 | |
| 9 | 66.7% | 62.0 | 2026-11-06 | |
| 10 | 64.4% | – | 2026-11-06 | |
| 11 | 63.3% | 60.7 | 2026-11-06 | |
| 12 | 63.3% | – | 2026-11-06 | |
| 13 | 63.3% | – | 2026-11-06 | |
| 14 | 61.1% | 64.4 | 2026-11-06 | |
| 15 | 61.1% | – | 2026-11-06 | |
| 16 | 58.9% | 61.8 | 2026-11-11 | |
| 17 | 58.9% | 58.7 | 2026-11-11 | |
| 18 | 58.1% | 57.9 | 2026-11-06 | |
| 19 | New | 56.7% | – | 2026-11-11 |
| 20 | 55.6% | – | 2026-11-11 | |
| 21 | 54.4% | – | 2026-11-06 | |
| 22 | 53.3% | 57.8 | 2026-11-11 | |
| 23 | 41.1% | – | 2026-11-06 | |
| 24 | 40.0% | 56.1 | 2026-11-11 | |
| 25 | 40.0% | 52.9 | 2026-11-11 | |
| 26 | 27.8% | 45.8 | 2026-11-11 | |
| 27 | 26.7% | 42.1 | 2026-11-11 | |
| 28 | 14.4% | 46.1 | 2026-11-11 |
AI coding benchmark leaderboard · All LLM benchmarks
Related benchmarks
Top head-to-heads on SWE Atlas v1.1: Test Writing
Cite as: BenchLeader, “SWE Atlas v1.1: Test Writing leaderboard”, data as of 16 Nov 2026, https://www.benchleader.com/benchmarks/scale_swe_atlas_tw.
SWE Atlas v1.1: Test Writing: questions
- What does SWE Atlas v1.1: Test Writing measure?
- Writing tests for existing code, SWE Atlas v1.1. Scale AI. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on SWE Atlas v1.1: Test Writing?
- As of 16 Nov 2026, the highest score on SWE Atlas v1.1: Test Writing is 78.9% by Claude Fable 5.1 (result dated 6 Nov 2026), ahead of Claude Fable 5.2 at 77.8%. Scores come from Scale AI SEAL and are re-read daily.
- Which AI model leads SWE Atlas v1.1: Test Writing?
- Claude Fable 5.1 leads SWE Atlas v1.1: Test Writing with 78.9% as of 16 Nov 2026 (result dated 6 Nov 2026), ahead of Claude Fable 5.2 at 77.8%.
- How many models have SWE Atlas v1.1: Test Writing results?
- 28 model configurations have a SWE Atlas v1.1: Test Writing result on BenchLeader, all taken from Scale AI SEAL.
- Who runs SWE Atlas v1.1: Test Writing and how often is it updated?
- SWE Atlas v1.1: Test Writing is published by Scale AI SEAL. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does SWE Atlas v1.1: Test Writing count toward the BenchLeader Index?
- No. SWE Atlas v1.1: Test Writing is shown for reference but left out of the composite index.