Skip to main content
BenchLM
Data

Artificial Analysis ITBench-AA (AA ITBench)

We show this table for reference; we do not rank on it.

Data verified

An independently evaluated IT-operations benchmark from Artificial Analysis.

Benchmark score on AA ITBench — November 13, 2026

We compile the AA ITBench rows from secondary reports. GPT-5.6 Sol leads the table at 56.2%, followed by Step 5 Preview (55.6%) and Gemini 3.8 Flash (52.5%). We do not use these results to rank models overall.

1Closed

GPT-5.6 Sol

OpenAI

Carried from an upstream page
56.2%
Overall 72.48Context 1.05M
2Open weights

Step 5 Preview

StepFun

Carried from an upstream page
55.6%
Overall 55.46Context 1M
3Closed

Gemini 3.8 Flash

Google

Carried from an upstream page
52.5%
Overall 64.47Context 1M

24 modelsAgenticCurrentDisplay onlyUpdated November 13, 2026

Benchmark score table (24 models)

None of the 24 results on this page has a documented run by the benchmark or an independent evaluator: 24 carried from an upstream page, where who ran them is not recorded.

Benchmark score results for AA ITBench
RankModel / configurationScoreParameters (B)Open / closed
1GPT-5.6 SolOpenAICarried from an upstream page
56.2%
Not reportedClosed
2Step 5 PreviewStepFunCarried from an upstream page
55.6%
Not reportedOpen
3Gemini 3.8 FlashGoogleCarried from an upstream page
52.5%
Not reportedClosed
4GLM-5.3-FlashZ.AICarried from an upstream page
51.2%
Not reportedOpen
5GPT-5.6 TerraOpenAICarried from an upstream page
51.0%
Not reportedClosed
6Claude Fable 5.1AnthropicCarried from an upstream page
49.5%
Not reportedClosed
7GPT-6 SolOpenAICarried from an upstream page
49.4%
Not reportedClosed
8GPT-6 AstraOpenAICarried from an upstream page
48.6%
Not reportedClosed
9Kimi K3Moonshot AICarried from an upstream page
47.7%
Not reportedOpen
10DeepSeek V4.1 Flash 1017DeepSeekCarried from an upstream page
46.9%
Not reportedOpen
11Claude Opus 4.7 (Adaptive)AnthropicCarried from an upstream page
46.7%
Not reportedClosed
12GLM-5.3Z.AICarried from an upstream page
46.1%
Not reportedOpen
13GPT-5.5OpenAICarried from an upstream page
45.8%
Not reportedClosed
14GLM-5.2Z.AICarried from an upstream page
42.7%
Not reportedOpen
15Qwen3.7 MaxAlibabaCarried from an upstream page
42.5%
Not reportedClosed
16Grok 4.7xAICarried from an upstream page
42.1%
Not reportedClosed
17Gemini 3.5 FlashGoogleCarried from an upstream page
40.3%
Not reportedClosed
18GPT-5.6 LunaOpenAICarried from an upstream page
40.3%
Not reportedClosed
19GLM-5.1Z.AICarried from an upstream page
40.3%
Not reportedOpen
20Claude Opus 5.5AnthropicCarried from an upstream page
38.2%
Not reportedClosed
21MiMo-V2.5-ProXiaomiCarried from an upstream page
38.2%
Not reportedClosed
22Gemma 4 31BGoogleCarried from an upstream page
37.3%
Not reportedOpen
23Qwen3.5-27BAlibabaCarried from an upstream page
35.5%
Not reportedOpen
24GPT-5.4 miniOpenAICarried from an upstream page
35.2%
Not reportedClosed
  • Lab-reported
  • Benchmark-run
  • Independent
What the labels mean

Among the reported AA ITBench rows, GPT-5.6 Sol is first at 56.2%. The third row is 3.7 points behind. The broader top-10 range is 9.3 points, so many of the published results sit in a relatively narrow band.

24 models have been evaluated on AA ITBench. The benchmark falls in the Agentic category. AA ITBench is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About AA ITBench

Year

2026

Tasks

IT incident-response tasks

Format

Task success rate

Difficulty

Enterprise IT operations

Stored as a display-only agentic row.

Artificial Analysis ITBench-AA Benchmark Leaderboard

Freshness and provenance

Version

AA ITBench 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does AA ITBench measure?

An independently evaluated IT-operations benchmark from Artificial Analysis.

Which model scores highest on AA ITBench?

GPT-5.6 Sol by OpenAI currently leads with a score of 56.2% on AA ITBench.

How many models are evaluated on AA ITBench?

24 AI models have published results on AA ITBench in the BenchLM catalog.

Compare top models on AA ITBench

GPT-5.6 Sol vs Step 5 PreviewStep 5 Preview vs Gemini 3.8 FlashGemini 3.8 Flash vs GLM-5.3-FlashGLM-5.3-Flash vs GPT-5.6 Terra
Last updated: November 13, 2026 · BenchLM version AA ITBench 2026

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 7,000+ subscribers.

One email each week. Unsubscribe anytime.