Noometry

Agentic & Tool Use benchmark

GDP.pdf leaderboard

As of October 2026, GPT-6 Astra has the highest published GDP.pdf score on Noometry at 34.2%, out of 36 models with results.

Last verified

About GDP.pdf

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. GPT-6 Astra 34.2%
  2. GPT-6.1 Sol 32%
  3. GPT-5.6 Sol 30.7%
  4. Claude Opus 5.5 30.6%
  5. Claude Fable 5 30%
  6. Claude Fable 5.1 29.6%
  7. Muse Spark 1.3 27.6%
  8. GPT-6 Sol 26.4%
  9. GPT-5.5 26%
  10. GPT-5.6 Terra 24.7%
  11. Claude Opus 4.8 24%
  12. Claude Opus 5 24%
  13. Gemini 3.7 Flash 23.8%
  14. Gemini 3.8 Flash 23.4%
  15. Qwen3.8 Max 23.2%

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

All results

#ModelProviderScoreSettingSourceDate
1GPT-6 Astra OpenAI34.2%maxEpoch AI
2GPT-6.1 Sol OpenAI32%highModel card (self-reported)2026-09-29
3GPT-5.6 Sol OpenAI30.7%Epoch AI
4Claude Opus 5.5 Anthropic30.6%maxEpoch AI
5Claude Fable 5 Anthropic30%Epoch AI
6Claude Fable 5.1 Anthropic29.6%highEpoch AI
7Muse Spark 1.3 Meta27.6%Epoch AI
8GPT-6 Sol OpenAI26.4%maxEpoch AI
9GPT-5.5 OpenAI26%xhighEpoch AI
10GPT-5.6 Terra OpenAI24.7%Epoch AI
11Claude Opus 4.8 Anthropic24%maxEpoch AI
12Claude Opus 5 Anthropic24%maxEpoch AI
13Gemini 3.7 Flash Google23.8%highEpoch AI
14Gemini 3.8 Flash Google23.4%mediumEpoch AI
15Qwen3.8 Max Alibaba (Qwen)23.2%xhighEpoch AI
16GPT-6 Luna OpenAI23%maxEpoch AI
17Grok 4.7 xAI22.8%xhighEpoch AI
18GPT-5.6 Luna OpenAI22.7%Epoch AI
19Claude Opus 4.7 Anthropic21%maxEpoch AI
20DeepSeek V4.1 Flash DeepSeek19.8%maxEpoch AI
21Kimi K3 Moonshot AI19%maxEpoch AI
22Claude Sonnet 4.6 Anthropic18%maxEpoch AI
23Grok 4.6 xAI17.2%xhighEpoch AI
24Gemini 3.1 Pro Preview Google17%Epoch AI
25Qwen3.8 Flash Alibaba (Qwen)16.6%xhighEpoch AI
26Muse Spark 1.2 Meta16%xhighEpoch AI
27Muse Spark 1.1 Meta15%Epoch AI
28Gemini 3.5 Flash Google14%Epoch AI
29Gemini 3.6 Flash Google14%Epoch AI
30GLM-5.3-Flash Z.ai (Zhipu)14%maxEpoch AI
31Grok 4.5 xAI14%highEpoch AI
32Kimi K2.6 Moonshot AI12%Epoch AI
33Gemini 3.5 Flash Lite Google10%highEpoch AI
34Gemini 3 Flash Preview Google10%Epoch AI
35Grok 4.3 xAI8%Epoch AI
36Nova 2.0 Pro Preview Amazon2%Epoch AI

Compare the leaders

Other agentic & tool use benchmarks

Frequently asked questions

Which model has the highest GDP.pdf score?

As of October 2026, GPT-6 Astra has the highest published GDP.pdf score on Noometry at 34.2%, out of 36 models with results.

What is the best open-weight model on GDP.pdf?

DeepSeek V4.1 Flash has the highest GDP.pdf accuracy among open-weight models at 19.8%, ranking 20 of 36 overall.