BenchLeader
Menu
Choose a modelLeaderboardsModelsBenchmarksProvidersReleases16

AA Cyber Index

Artificial Analysis' composite of its three security evaluations. Shown for orientation; the boards beneath it are what the index uses.

As of 18 Oct 2026, Grok 4.7 leads AA Cyber Index on BenchLeader with 5643.2%, ahead of MiMo-V2.6-Pro at 5606.6%, across 18 model configurations with a published result.

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

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

Top 15
  1. 1Grok 4.7 (xhigh)5643.2%
  2. 2MiMo-V2.6-Pro5606.6%
  3. 3GPT-6 Luna (max)5268.5%
  4. 4GLM 5.3 Flash4992.4%
  5. 5Mistral Large 44951.1%
  6. 6Step 5 Preview4770.7%
  7. 7Kimi K3 (max)4126.8%
  8. 8DeepSeek V4.1 Flash (max)4081.1%
  9. 9GPT-6 Sol (max)3682.8%
  10. 10GLM 5.3 (max)3637.6%
  11. 11Beam 1.03369.9%
  12. 12GPT-6 Astra (max)3342.0%
  13. 13Claude Opus 5.5 (max)2866.4%
  14. 14Gemini 3.8 Flash (high)2514.5%
  15. 15Claude Fable 5.1 (max)2501.9%
18 of 18
#
1Grok 4.7xhigh5643.2%60.62026-09-21
2MiMo-V2.6-Pro5606.6%65.32026-09-21
3GPT-6 Lunamax5268.5%58.62026-09-22
4GLM 5.3 Flash4992.4%62.22026-08-26
5Mistral Large 44951.1%58.72026-10-06
6Step 5 Preview4770.7%62.22026-09-18
7Kimi K3max4126.8%64.82026-07-16
8DeepSeek V4.1 Flashmax4081.1%61.52026-09-10
9GPT-6 Solmax3682.8%65.62026-09-22
10GLM 5.3max3637.6%63.62026-08-18
11NewBeam 1.03369.9%56.42026-10-16
12GPT-6 Astramax3342.0%69.72026-09-03
13Claude Opus 5.5max2866.4%71.62026-09-22
14Gemini 3.8 Flashhigh2514.5%63.32026-09-02
15Claude Fable 5.1max2501.9%69.92026-09-01
16MiniMax M32238.8%56.22026-06-01
17Nemotron 3 Ultra 550B A55Bthinking1307.8%55.62026-06-04
18Qwen3.8 27Bxhigh1272.4%56.32026-08-14

More composite benchmarks · All LLM benchmarks

Cite as: BenchLeader, “AA Cyber Index leaderboard”, https://www.benchleader.com/benchmarks/aa_cyber_index, data as of 18 Oct 2026.

AA Cyber Index: questions

What does AA Cyber Index measure?
Artificial Analysis' composite of its three security evaluations. Shown for orientation; the boards beneath it are what the index uses. Scores are reported in percent of tasks solved; higher is better.
Which AI model leads AA Cyber Index?
Grok 4.7 leads AA Cyber Index with 5643.2% as of 18 Oct 2026, ahead of MiMo-V2.6-Pro at 5606.6%.
How many models have AA Cyber Index results?
18 model configurations have a AA Cyber Index result on BenchLeader, all taken from Artificial Analysis.
Who runs AA Cyber Index and how often is it updated?
AA Cyber Index is published by Artificial Analysis. BenchLeader re-reads the published results once a day and records the date each result was published.
Does AA Cyber Index count toward the BenchLeader Index?
No. AA Cyber Index is shown for reference but left out of the composite index.