BenchLeader

SWE Atlas v1.1: Refactoring leaderboard

As of 16 Nov 2026, the top score on SWE Atlas v1.1: Refactoring is 80.0% by Claude Fable 5.2 (Anthropic), ahead of Claude Opus 5.5 at 77.1%, across 28 model configurations with a published result (Scale AI SEAL, result dated 11 Nov 2026). Refactoring tasks in real repositories, SWE Atlas v1.1. Scale AI. SWE Atlas v1.1: Refactoring is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 16 Nov 2026)

  • Claude Fable 5.2 holds the top SWE Atlas v1.1: Refactoring score at 80.0% (Scale AI SEAL, 11 Nov 2026).
  • Claude Opus 5.5 is second on SWE Atlas v1.1: Refactoring at 77.1% (Scale AI SEAL, 6 Nov 2026).
  • SWE Atlas v1.1: Refactoring: Refactoring tasks in real repositories, SWE Atlas v1.1. Scale AI (Scale AI SEAL).
  • 28 model configurations have a published SWE Atlas v1.1: Refactoring result on BenchLeader (BenchLeader, 16 Nov 2026).
  • SWE Atlas v1.1: Refactoring does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

Refactoring tasks in real repositories, 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: Refactoring.

28 of 28
#
1NewClaude Fable 5.280.0%65.42026-11-11
2Claude Opus 5.577.1%61.02026-11-06
3GPT-6 Astra71.4%66.42026-11-06
4Claude Fable 5.171.4%66.42026-11-06
5Gemini 4 Argon67.1%64.42026-11-06
6Gemini 4 Carbon64.3%–2026-11-11
7Claude Sonnet 5.562.9%–2026-11-06
8Muse Wonder60.0%64.42026-11-06
9GPT-6.1 Sol60.0%62.02026-11-06
10Gemini 3.8 Flash55.7%56.42026-11-06
11GPT-6.1 Terra55.7%–2026-11-06
12Kimi K3.1 Pro54.3%61.82026-11-11
13Kimi K354.3%60.72026-11-06
14DeepSeek V4.1 Flash54.3%–2026-11-06
15Inkling 2 Small53.3%57.92026-11-06
16Qwen3.8 Maxmax52.9%58.72026-11-11
17Claude Haiku 5.552.9%–2026-11-06
18NewMuse Ember50.0%–2026-11-11
19GLM 5.348.6%–2026-11-11
20Grok 4.748.6%–2026-11-06
21GPT-6 Luna47.1%–2026-11-06
22GLM 5.3 Flash45.7%57.82026-11-11
23Muse Spark Open35.7%–2026-11-06
24Mistral Large 434.3%52.92026-11-11
25MiniMax M328.6%56.12026-11-11
26Nemotron 3.5 Super VL20.0%45.82026-11-11
27Nemotron 3 Ultra15.7%42.12026-11-11
28gpt-oss-120b1.4%46.12026-11-11

AI coding benchmark leaderboard · All LLM benchmarks

Cite as: BenchLeader, “SWE Atlas v1.1: Refactoring leaderboard”, data as of 16 Nov 2026, https://www.benchleader.com/benchmarks/scale_swe_atlas_refactoring.

SWE Atlas v1.1: Refactoring: questions

What does SWE Atlas v1.1: Refactoring measure?
Refactoring tasks in real repositories, 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: Refactoring?
As of 16 Nov 2026, the highest score on SWE Atlas v1.1: Refactoring is 80.0% by Claude Fable 5.2 (result dated 11 Nov 2026), ahead of Claude Opus 5.5 at 77.1%. Scores come from Scale AI SEAL and are re-read daily.
Which AI model leads SWE Atlas v1.1: Refactoring?
Claude Fable 5.2 leads SWE Atlas v1.1: Refactoring with 80.0% as of 16 Nov 2026 (result dated 11 Nov 2026), ahead of Claude Opus 5.5 at 77.1%.
How many models have SWE Atlas v1.1: Refactoring results?
28 model configurations have a SWE Atlas v1.1: Refactoring result on BenchLeader, all taken from Scale AI SEAL.
Who runs SWE Atlas v1.1: Refactoring and how often is it updated?
SWE Atlas v1.1: Refactoring 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: Refactoring count toward the BenchLeader Index?
No. SWE Atlas v1.1: Refactoring is shown for reference but left out of the composite index.