BenchLeader
Menu
Choose a modelLeaderboardsModelsBenchmarksProvidersReleases16

FrontierCode 1.1 (AA)

Artificial Analysis' run of Cognition's FrontierCode. Epoch's run of the same benchmark is the one in the index.

As of 18 Oct 2026, Claude Opus 5.5 leads FrontierCode 1.1 (AA) on BenchLeader with 57.0%, ahead of Claude Opus 5.5 at 54.0%, across 30 model configurations with a published result.

Published by
Artificial Analysis
Category
Coding
Index weight
Reference only
Models
30
Data as of
18 Oct 2026

Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.

Top 15
  1. 1Claude Opus 5.5 (medium)57.0%
  2. 2Claude Opus 5.5 (high)54.0%
  3. 3Claude Opus 5.5 (xhigh)53.5%
  4. 4Claude Opus 5.5 (max)52.2%
  5. 5GPT-6.1 Sol (max)49.1%
  6. 6Claude Fable 5.1 (max)49.0%
  7. 7GPT-6 Astra (max)47.0%
  8. 8Claude Sonnet 5.5 (xhigh)46.1%
  9. 9Claude Sonnet 5.5 (max)45.8%
  10. 10Claude Haiku 5.5 (xhigh)45.5%
  11. 11MiMo-V2.6-Pro45.4%
  12. 12Claude Sonnet 5.5 (high)44.3%
  13. 13DeepSeek V4.1 Flash (max)44.0%
  14. 14Claude Haiku 5.5 (high)43.6%
  15. 15Kimi K3 (max)43.6%
30 of 30
#
1Claude Opus 5.5medium57.0%66.82026-09-22
2Claude Opus 5.5high54.0%69.72026-09-22
3Claude Opus 5.5xhigh53.5%69.62026-09-22
4Claude Opus 5.5max52.2%71.92026-09-22
5GPT-6.1 Solmax49.1%69.32026-09-29
6Claude Fable 5.1max49.0%70.12026-09-01
7GPT-6 Astramax47.0%69.92026-09-03
8Claude Sonnet 5.5xhigh46.1%66.92026-09-28
9Claude Sonnet 5.5max45.8%67.52026-09-28
10Claude Haiku 5.5xhigh45.5%59.92026-10-07
11MiMo-V2.6-Pro45.4%65.62026-09-21
12Claude Sonnet 5.5high44.3%66.12026-09-28
13DeepSeek V4.1 Flashmax44.0%61.72026-09-10
14Claude Haiku 5.5high43.6%59.52026-10-07
15Kimi K3max43.6%65.12026-07-16
16Gemini 4 Argonhigh43.5%71.42026-09-30
17Qwen3.8 Maxmax43.2%63.92026-09-02
18GPT-6 Lunamax41.9%58.72026-09-22
19Claude Haiku 5.5max41.8%57.92026-10-07
20GLM 5.3 Flash41.0%62.42026-08-26
21Muse Spark 1.3max40.6%67.42026-09-02
22GLM 5.3max40.1%63.82026-08-18
23Step 5 Preview38.4%62.42026-09-18
24Gemini 3.8 Flashhigh36.4%63.52026-09-02
25Qwen3.8 27Bxhigh34.2%56.52026-08-14
26Grok 4.7xhigh33.8%60.72026-09-21
27K2 Horizon 375B A23B32.1%56.62026-09-03
28Claude Haiku 5.5low29.3%55.02026-10-07
29MiniMax M323.0%56.42026-06-01
30Nemotron 3 Ultra 550B A55Bthinking10.5%55.92026-06-04

More coding benchmarks · All LLM benchmarks

Cite as: BenchLeader, “FrontierCode 1.1 (AA) leaderboard”, https://www.benchleader.com/benchmarks/aa_frontiercode, data as of 18 Oct 2026.

FrontierCode 1.1 (AA): questions

What does FrontierCode 1.1 (AA) measure?
Artificial Analysis' run of Cognition's FrontierCode. Epoch's run of the same benchmark is the one in the index. Scores are reported in percent of tasks solved; higher is better.
Which AI model leads FrontierCode 1.1 (AA)?
Claude Opus 5.5 leads FrontierCode 1.1 (AA) with 57.0% as of 18 Oct 2026, ahead of Claude Opus 5.5 at 54.0%.
How many models have FrontierCode 1.1 (AA) results?
30 model configurations have a FrontierCode 1.1 (AA) result on BenchLeader, all taken from Artificial Analysis.
Who runs FrontierCode 1.1 (AA) and how often is it updated?
FrontierCode 1.1 (AA) is published by Artificial Analysis. BenchLeader re-reads the published results once a day and records the date each result was published.
Does FrontierCode 1.1 (AA) count toward the BenchLeader Index?
No. FrontierCode 1.1 (AA) is shown for reference but left out of the composite index.