Benchmark Almanac

sudo L7

Claude Fable 5.2 52.2% leads since November 2026
In use

Coding agents get staff-level engineering requests in mostly private production codebases. Graders score whether the agent found hidden risks and shipped the right change.

60 tasks. Each model runs in its own developer's coding agent.

Made by
Surge AI
Released
October 2026
Kind
Fixing real repositories
Level
Expert
Scored
Graded by people
Answers
Agent task
Tasks
Kept private
Strongest model tested
Claude Fable 5.2 #1 on the capability index
Last in a launch post
November 2026
Last model added
November 2026
Scores refreshed
November 12th, 2026
19
Entries on its own leaderboard, 18 models we track
1
Launch post lists it, from 1 lab

Claude Fable 5.2 leads with 52.2%

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

19 entries. Scroll sideways for the rest.

  1. 52.2 Claude Fable 5.2
  2. 50.6 Claude Sonnet 5.5
  3. 46.1 Claude Opus 5.5
  4. 33.9 GPT-6.1 Sol
  5. 32.2 Claude Fable 5.1
  6. 31.1 GPT-6 Astra
  7. 18.9 MiniMax-M3.1-Flash
  8. 16.7 Qwen 3.8 Max
  9. 16.7 Kimi K3.1 Pro
  10. 16.1 Muse Wonder 1.0
  11. 16.1 Gemini 4 Argon
  12. 15.6 Tencent Hy4 preview
  13. 14.1 Inkling 2 Small
  14. 13.9 Mistral Large 4
  15. 12.8 Muse Spark 1.3
  16. 12.2 Kimi K3
  17. 12.2 Grok 4.7
  18. 10.6 Gemini 3.8 Flash
  19. 9.40 GLM-5.3
  1. 1 Claude Fable 5.2 adaptive/max 52.2%
  2. 2 Claude Sonnet 5.5 adaptive/max 50.6%
  3. 3 Claude Opus 5.5 adaptive/max 46.1%
  4. 4 GPT-6.1 Sol max 33.9%
  5. 5 Claude Fable 5.1 adaptive/max 32.2%
  6. 6 GPT-6 Astra max 31.1%
  7. 7 MiniMax-M3.1-Flash max 18.9%
  8. 8 Qwen 3.8 Max xhigh 16.7%
  9. 9 Kimi K3.1 Pro max 16.7%
  10. 10 Muse Wonder 1.0 max 16.1%
  11. 11 Gemini 4 Argon high 16.1%
  12. 12 Tencent Hy4 preview high 15.6%
  13. 13 Inkling 2 Small xhigh 14.1%
  14. 14 Mistral Large 4 high 13.9%
  15. 15 Muse Spark 1.3 max 12.8%
  16. 16 Kimi K3 max 12.2%
  17. 17 Grok 4.7 xhigh 12.2%
  18. 18 Gemini 3.8 Flash high 10.6%
  19. 19 GLM-5.3 max 9.40%

Source: Official sudo L7 leaderboard

#EntryScoreSettingDateSource
1 Claude Fable 5.2 52.2% adaptive/max Nov 2026 Official leaderboard
2 Claude Sonnet 5.5 50.6% adaptive/max Sep 2026 Official leaderboard
3 Claude Opus 5.5 46.1% adaptive/max Sep 2026 Official leaderboard
4 GPT-6.1 Sol 33.9% max Sep 2026 Official leaderboard
5 Claude Fable 5.1 32.2% adaptive/max Sep 2026 Official leaderboard
6 GPT-6 Astra 31.1% max Sep 2026 Official leaderboard
7 MiniMax-M3.1-Flash 18.9% max Oct 2026 Official leaderboard
8 Qwen 3.8 Max 16.7% xhigh Aug 2026 Official leaderboard
9 Kimi K3.1 Pro 16.7% max Oct 2026 Official leaderboard
10 Muse Wonder 1.0 16.1% max Official leaderboard
11 Gemini 4 Argon 16.1% high Sep 2026 Official leaderboard
12 Tencent Hy4 preview 15.6% high Aug 2026 Official leaderboard
13 Inkling 2 Small 14.1% xhigh Nov 2026 Official leaderboard
14 Mistral Large 4 13.9% high Oct 2026 Official leaderboard
15 Muse Spark 1.3 12.8% max Sep 2026 Official leaderboard
16 Kimi K3 12.2% max Jul 2026 Official leaderboard
17 Grok 4.7 12.2% xhigh Sep 2026 Official leaderboard
18 Gemini 3.8 Flash 10.6% high Sep 2026 Official leaderboard
19 GLM-5.3 9.40% max Aug 2026 Official leaderboard

Labs report scores for 1 model in their cards. We plot every source in the charts below.

2 models have held number one on the official leaderboard since its release in October 2026

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard 2 leaders
Claude Fable 5.2 52.2%

Source: Official sudo L7 leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Sonnet 5.5 Claude Opus 5.5 Claude Fable 5.1 Qwen 3.8 Max Kimi K3 0% 20% 40% 60% 80% 100% 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Fable 5.1

Sources: Official sudo L7 leaderboard and labs’ model cards

One lab put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Thinking Machines
DateLaunch postScoreAs the post writes it
Nov 5, 2026 Inkling 2 Small 14.1 Sudo L7, 14.1%

One lab claim next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 1 of them match, which usually means one side copied the other. Every benchmark.

Model Lab says Board says Gap
Inkling 2 Small 14.1% Sudo L7, 14.1% 14.1% Leaderboard, xhigh Same

More in Coding

All 282 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471