BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases7

GDP.pdf leaderboard

As of 6 Nov 2026, the top score on GDP.pdf is 35.0% by GPT-6.1 Sol (OpenAI), ahead of GPT-6 Astra at 34.2%, across 56 model configurations with a published result (Epoch AI Benchmarking Hub, result dated 29 Sep 2026). Economically valuable tasks that hinge on reading and producing PDF documents. GDP.pdf is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 6 Nov 2026)

  • GPT-6.1 Sol holds the top GDP.pdf score at 35.0% (Epoch AI Benchmarking Hub, 29 Sep 2026).
  • GPT-6 Astra is second on GDP.pdf at 34.2% (Epoch AI Benchmarking Hub, 3 Sep 2026).
  • GDP.pdf: Professional document tasks over PDFs (Epoch AI Benchmarking Hub).
  • 56 model configurations have a published GDP.pdf result on BenchLeader (BenchLeader, 6 Nov 2026).
  • GDP.pdf does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

Professional document tasks over PDFs.

Published by
Epoch AI Benchmarking Hub
Category
Knowledge
Index weight
Reference only
Models
56
Data as of
6 Nov 2026
Newest result
12 Oct 2026

CC BY 4.0 — Epoch AI, ‘AI Benchmarking Hub’, epoch.ai/benchmarks. Mirrored boards credit their original publishers.

The same test, run by another publisher: GDP.pdf (AA).

What the test looks like

Economically valuable tasks that hinge on reading and producing PDF documents.

How it is scored

Score, as published.

What to keep in mind

New; small set.

Top 15
  1. 1GPT-6.1 Sol (max)35.0%
  2. 2GPT-6 Astra (max)34.2%
  3. 3Claude Sonnet 5.5 (max)31.4%
  4. 4Muse Wonder (max)30.8%
  5. 5GPT-5.6 Sol (max)30.7%
  6. 6GPT-5.6 Sol30.7%
  7. 7Claude Opus 5.5 (max)30.6%
  8. 8Claude Fable 530.0%
  9. 9Claude Fable 5 (max)29.8%
  10. 10Claude Fable 5.1 (high)29.6%
  11. 11Gemini 4 Argon (high)28.4%
  12. 12Claude Fable 5.1 (max)27.6%
  13. 13Muse Spark 1.3 (xhigh)27.6%
  14. 14Muse Spark 1.327.6%
  15. 15GPT-6 Sol (max)26.4%
56 of 56
#
1GPT-6.1 Solmax35.0%67.32026-09-29
2GPT-6 Astramax34.2%67.82026-09-03
3Claude Sonnet 5.5max31.4%65.42026-09-28
4Muse Wondermax30.8%62.32026-10-12
5GPT-5.6 Solmax30.7%65.42026-07-09
6GPT-5.6 Sol30.7%55.02026-07-09
7Claude Opus 5.5max30.6%69.32026-09-22
8Claude Fable 530.0%66.52026-06-09
9Claude Fable 5max29.8%65.32026-06-09
10Claude Fable 5.1high29.6%64.02026-09-01
11Gemini 4 Argonhigh28.4%66.02026-09-30
12Claude Fable 5.1max27.6%67.22026-09-01
13Muse Spark 1.3xhigh27.6%58.52026-09-02
14Muse Spark 1.327.6%57.72026-09-02
15GPT-6 Solmax26.4%63.32026-09-22
16GPT-5.5xhigh26.0%64.32026-04-23
17GPT-5.6 Terramedium24.7%52.92026-07-09
18GPT-5.6 Terra24.7%51.52026-07-09
19Claude Opus 5max24.0%65.42026-07-24
20Claude Opus 4.8max24.0%60.22026-05-28
21Gemini 3.7 Flashhigh23.8%62.22026-08-13
22Gemini 3.8 Flashmedium23.4%60.72026-09-02
23Gemini 3.8 Flashhigh23.2%62.02026-09-02
24Qwen3.8 Maxxhigh23.2%57.12026-08-02
25GPT-6 Lunamax23.0%56.72026-09-22
26Grok 4.7xhigh22.8%58.32026-09-21
27GPT-5.6 Lunamedium22.7%50.72026-07-09
28GPT-5.6 Luna22.7%47.62026-07-09
29Gemini 3.7 Flashmedium21.8%60.32026-08-13
30Claude Opus 4.7max21.0%59.52026-04-16
31DeepSeek V4.1 Flashmax19.8%59.02026-09-09
32Kimi K3max19.0%62.12026-07-16
33Claude Sonnet 4.6max18.0%54.22026-02-17
34Grok 4.6xhigh17.2%57.92026-08-12
35Gemini 3.1 Pro17.0%60.52026-02-19
36Gemini 3.1 Prohigh17.0%60.12026-02-19
37Qwen3.8 Flashxhigh16.6%–2026-08-26
38Muse Spark 1.2xhigh16.0%61.72026-08-05
39Grok 4.6high16.0%60.52026-08-12
40Muse Spark 1.115.0%62.22026-07-09
41Muse Spark 1.1medium15.0%–2026-07-09
42Gemini 3.5 Flashmedium14.0%61.92026-05-19
43Gemini 3.6 Flashhigh14.0%61.22026-07-21
44Grok 4.5high14.0%56.92026-07-08
45GLM 5.3 Flashmax14.0%54.22026-08-20
46Gemini 3.6 Flash14.0%49.52026-07-21
47Gemini 3.5 Flash14.0%48.62026-05-19
48Kimi K2.612.0%55.22026-04-20
49Muse Spark 1.212.0%–2026-08-05
50Gemini 3 Flashhigh10.0%58.32025-12-17
51Gemini 3 Flash10.0%56.92025-12-17
52Gemini 3.5 Flash Litehigh10.0%47.92026-07-21
53Grok 4.3high8.0%56.62026-04-17
54Grok 4.38.0%54.52026-04-17
55Nova 2.0 Pro Previewno reasoning2.0%–2025-12-02
56Nova 2.0 Pro Preview2.0%–2025-12-02

More knowledge benchmarks · All LLM benchmarks

Related benchmarks

Top head-to-heads on GDP.pdf

Cite as: BenchLeader, “GDP.pdf leaderboard”, data as of 6 Nov 2026, https://www.benchleader.com/benchmarks/gdp_pdf.

GDP.pdf: questions

What does GDP.pdf measure?
Economically valuable tasks that hinge on reading and producing PDF documents. Scores are reported in percent of tasks solved; higher is better.
What is the top score on GDP.pdf?
As of 6 Nov 2026, the highest score on GDP.pdf is 35.0% by GPT-6.1 Sol (result dated 29 Sep 2026), ahead of GPT-6 Astra at 34.2%. Scores come from Epoch AI Benchmarking Hub and are re-read daily.
Which AI model leads GDP.pdf?
GPT-6.1 Sol leads GDP.pdf with 35.0% as of 6 Nov 2026 (result dated 29 Sep 2026), ahead of GPT-6 Astra at 34.2%.
How many models have GDP.pdf results?
56 model configurations have a GDP.pdf result on BenchLeader, all taken from Epoch AI Benchmarking Hub.
Who runs GDP.pdf and how often is it updated?
GDP.pdf is published by Epoch AI Benchmarking Hub. BenchLeader re-reads the published results once a day and records the date each result was published.
Does GDP.pdf count toward the BenchLeader Index?
No. GDP.pdf is shown for reference but left out of the composite index.