Noometry

Agentic & Tool Use benchmark

APEX-Agents leaderboard

As of October 2026, Gemini 4 Argon has the highest published APEX-Agents score on Noometry at 82.2%, out of 49 models with results.

Last verified

About APEX-Agents

A description with primary sources is being prepared for this benchmark.

Category
Agentic & Tool Use
Introduced
2026
Unit
Percent (random guessing ≈ 0%)
Official site
epoch.ai

Top 15 models

  1. Gemini 4 Argon 82.2%
  2. Claude Sonnet 5.5 75.5%
  3. Claude Opus 5.5 73.5%
  4. Claude Fable 5.1 68.6%
  5. Gemini 3.7 Flash 67.8%
  6. Claude Opus 5 65.8%
  7. Grok 4.6 65.3%
  8. GPT-6 Astra 64.7%
  9. Gemini 3.8 Flash 64.3%
  10. Claude Fable 5 63.6%
  11. Qwen3.8 Max 63.3%
  12. GPT-6.1 Sol 60%
  13. MiMo-V2.6-Pro 59.5%
  14. GPT-5.6 Terra 58.2%
  15. Muse Spark 1.3 57.8%

Sponsored placements are available on pages like this one. Advertise on Noometry

All results

#ModelProviderScoreSettingSourceDate
1Gemini 4 Argon Google82.2%Epoch AI
2Claude Sonnet 5.5 Anthropic75.5%maxEpoch AI
3Claude Opus 5.5 Anthropic73.5%maxEpoch AI
4Claude Fable 5.1 Anthropic68.6%Epoch AI
5Gemini 3.7 Flash Google67.8%Epoch AI
6Claude Opus 5 Anthropic65.8%maxEpoch AI
7Grok 4.6 xAI65.3%Epoch AI
8GPT-6 Astra OpenAI64.7%Epoch AI
9Gemini 3.8 Flash Google64.3%Epoch AI
10Claude Fable 5 Anthropic63.6%Epoch AI
11Qwen3.8 Max Alibaba (Qwen)63.3%Epoch AI
12GPT-6.1 Sol OpenAI60%maxEpoch AI
13MiMo-V2.6-Pro Xiaomi59.5%Epoch AI
14GPT-5.6 Terra OpenAI58.2%maxEpoch AI
15Muse Spark 1.3 Meta57.8%Epoch AI
16GLM-5.3 Z.ai (Zhipu)56.6%Epoch AI
17Grok 4.5 xAI56.2%Epoch AI
18GPT-5.5 OpenAI55.1%Epoch AI
19Grok 4.7 xAI54.6%Epoch AI
20Claude Sonnet 5 Anthropic54.5%Epoch AI
21GPT-6 Sol OpenAI54.3%maxEpoch AI
22GLM-5.3-Flash Z.ai (Zhipu)52.8%Epoch AI
23GPT-5.4 OpenAI52.4%Epoch AI
24GPT-5.6 Sol OpenAI51.4%promaxEpoch AI
25Kimi K3 Moonshot AI50.6%Epoch AI
26Claude Opus 4.7 Anthropic49.2%maxEpoch AI
27Claude Opus 4.8 Anthropic48.9%maxEpoch AI
28Qwen3.8 27B Alibaba (Qwen)47.5%Epoch AI
29DeepSeek V4 Pro DeepSeek47.3%Epoch AI
30Gemini 3.6 Flash Google46.9%Epoch AI
31Claude Opus 4.6 Anthropic46.3%maxEpoch AI
32GLM-5.2 Z.ai (Zhipu)45.2%Epoch AI
33GPT-6 Luna OpenAI44.3%maxEpoch AI
34Claude Sonnet 4.6 Anthropic43%highEpoch AI
35GPT-5.6 Luna OpenAI43%Epoch AI
36GLM-5.1 Z.ai (Zhipu)40.9%Epoch AI
37DeepSeek V4.1 Flash DeepSeek39.5%Epoch AI
38MiniMax-M3 MiniMax37.7%Epoch AI
39Kimi K2.7 Code Moonshot AI37.6%Epoch AI
40Muse Spark 1.2 Meta36.4%Epoch AI
41Gemini 3.1 Pro Preview Google35.3%Epoch AI
42Inkling Thinking Machines Lab33.8%Epoch AI
43Muse Spark 1.1 Meta31.8%Epoch AI
44Gemini 3.5 Flash Lite Google29.3%Epoch AI
45Gemini 3.5 Flash Google27.5%Epoch AI
46Qwen3.5 397B-A17B Alibaba (Qwen)24.9%Epoch AI
47Nemotron 3 Ultra NVIDIA22.7%Epoch AI
48DeepSeek-V3.2-Exp DeepSeek21.3%Epoch AI
49gpt-oss-120b OpenAI4.4%Epoch AI

Compare the leaders

Other agentic & tool use benchmarks

Frequently asked questions

Which model has the highest APEX-Agents score?

As of October 2026, Gemini 4 Argon has the highest published APEX-Agents score on Noometry at 82.2%, out of 49 models with results.

What is the best open-weight model on APEX-Agents?

MiMo-V2.6-Pro has the highest APEX-Agents accuracy among open-weight models at 59.5%, ranking 13 of 49 overall.