Noometry

Agentic & Tool Use benchmark

METR Time Horizons leaderboard

As of October 2026, Claude Mythos Preview has the highest published METR Time Horizons score on Noometry at 85.2%, out of 32 models with results.

Last verified

About METR Time Horizons

METR's measure of how long a task (in human expert time) a model can complete with 50% reliability.

Category
Agentic & Tool Use
Introduced
2025
Format
Software tasks
Unit
Percent (random guessing ≈ 0%)
Official site
metr.org

Top 15 models

  1. Claude Mythos Preview 85.2%
  2. Claude Opus 4.6 78.9%
  3. Gemini 3.1 Pro Preview 77%
  4. GPT-5.2 75.3%
  5. Claude Opus 4.5 75%
  6. GPT-5.3 Codex 74.5%
  7. GPT-5.4 74.3%
  8. Gemini 3 Pro 71%
  9. GPT-5.1-Codex 70.8%
  10. GPT-5 69.6%
  11. Claude Sonnet 4.5 67.4%
  12. Claude Opus 4.1 66.8%
  13. Grok 4 66.6%
  14. o3 65.4%
  15. Claude Opus 4 63.9%

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

All results

#ModelProviderScoreSettingSourceDate
1Claude Mythos Preview Anthropic85.2%Epoch AI
2Claude Opus 4.6 Anthropic78.9%Epoch AI
3Gemini 3.1 Pro Preview Google77%Epoch AI
4GPT-5.2 OpenAI75.3%highEpoch AI
5Claude Opus 4.5 Anthropic75%16KEpoch AI
6GPT-5.3 Codex OpenAI74.5%Epoch AI
7GPT-5.4 OpenAI74.3%xhighEpoch AI
8Gemini 3 Pro Google71%Epoch AI
9GPT-5.1-Codex OpenAI70.8%Epoch AI
10GPT-5 OpenAI69.6%mediumEpoch AI
11Claude Sonnet 4.5 Anthropic67.4%16KEpoch AI
12Claude Opus 4.1 Anthropic66.8%16KEpoch AI
13Grok 4 xAI66.6%Epoch AI
14o3 OpenAI65.4%mediumEpoch AI
15Claude Opus 4 Anthropic63.9%16KEpoch AI
16o4-mini OpenAI63.9%mediumEpoch AI
17Claude Sonnet 4 Anthropic62%16KEpoch AI
18Claude 3.7 Sonnet Anthropic60%16KEpoch AI
19Kimi K2 (Jul 2025) Moonshot AI59.2%Epoch AI
20gpt-oss-120b OpenAI56.6%Epoch AI
21Gemini 2.5 Pro Google55.4%Epoch AI
22DeepSeek-R1 DeepSeek53.8%Epoch AI
23o1 OpenAI51.1%mediumEpoch AI
24DeepSeek-V3 DeepSeek49.6%Epoch AI
25Claude 3.5 Sonnet Anthropic45.2%Epoch AI
26GPT-4o OpenAI40.8%Epoch AI
27GPT-4 Turbo OpenAI36.7%Epoch AI
28GPT-4 OpenAI36.1%Epoch AI
29Qwen2.5 72B Instruct Alibaba (Qwen)35.8%Epoch AI
30Qwen2-72B Alibaba (Qwen)29.9%Epoch AI
31Claude 3 Opus Anthropic29.5%Epoch AI
32GPT-3.5-turbo OpenAI21.5%Epoch AI

Compare the leaders

Other agentic & tool use benchmarks

Frequently asked questions

What does METR Time Horizons measure?

METR's measure of how long a task (in human expert time) a model can complete with 50% reliability.

Which model has the highest METR Time Horizons score?

As of October 2026, Claude Mythos Preview has the highest published METR Time Horizons score on Noometry at 85.2%, out of 32 models with results.

What is the best open-weight model on METR Time Horizons?

Kimi K2 (Jul 2025) has the highest METR Time Horizons accuracy among open-weight models at 59.2%, ranking 19 of 32 overall.