GDP.pdf (AA) leaderboard
As of 9 Nov 2026, the top score on GDP.pdf (AA) is 32.6% by Muse Wonder Full (Meta), ahead of GPT-6 Astra at 32.2%, across 252 model configurations with a published result (Artificial Analysis, result dated 12 Oct 2026). Document work on real PDFs, scored all-or-nothing: every required element has to be right for the task to pass, which is why the scores look low. GDP.pdf (AA) is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 9 Nov 2026)
- Muse Wonder Full holds the top GDP.pdf (AA) score at 32.6% (Artificial Analysis, 12 Oct 2026).
- GPT-6 Astra is second on GDP.pdf (AA) at 32.2% (Artificial Analysis, 3 Sep 2026).
- GDP.pdf (AA): Document work on real PDFs, scored all-or-nothing: every required element has to be right for the task to pass, which is why the scores look low (Artificial Analysis).
- 252 model configurations have a published GDP.pdf (AA) result on BenchLeader (BenchLeader, 9 Nov 2026).
- GDP.pdf (AA) does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Document work on real PDFs, scored all-or-nothing: every required element has to be right for the task to pass, which is why the scores look low.
- Published by
- Artificial Analysis
- Category
- Agents & tools
- Index weight
- Reference only
- Models
- 252
- Data as of
- 9 Nov 2026
Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.
The same test, run by another publisher: GDP.pdf.
- 1Muse Wonder Full (max)32.6%
- 2GPT-6 Astra (xhigh)32.2%
- 3GPT-6.1 Sol (high)32.0%
- 4GPT-6.1 Sol (xhigh)31.8%
- 5Kimi K3.1 Pro31.4%
- 6GPT-6 Astra (max)31.0%
- 7GPT-6.1 Sol (max)31.0%
- 8GPT-6 Astra (high)31.0%
- 9Muse Wonder (max)30.6%
- 10GPT-6 Astra (medium)30.4%
- 11GPT-6 Astra (low)30.4%
- 12GPT-6.1 Sol (medium)30.0%
- 13Gemini 4 Carbon (medium)29.6%
- 14Muse Wonder (high)29.2%
- 15Claude Opus 5.5 (high)28.8%
252 of 252
| # | ||||||
|---|---|---|---|---|---|---|
| 1 |  Muse Wonder Fullmax | 32.6% | 62.2 | – | 12k | 2026-10-12 |
| 2 |  GPT-6 Astraxhigh | 32.2% | 66.2 | $1.91 | 6k | 2026-09-03 |
| 3 | GPT-6.1 Solhigh | 32.0% | 63.5 | $0.349 | 3k | 2026-09-29 |
| 4 | GPT-6.1 Solxhigh | 31.8% | 63.8 | $0.368 | 5k | 2026-09-29 |
| 5 |  Kimi K3.1 Proopen ↗ | 31.4% | 61.2 | $0.522 | 19k | 2026-10-23 |
| 6 | GPT-6 Astramax | 31.0% | 67.4 | $2.08 | 9k | 2026-09-03 |
| 7 | GPT-6.1 Solmax | 31.0% | 67.0 | $0.420 | 10k | 2026-09-29 |
| 8 | GPT-6 Astrahigh | 31.0% | 66.5 | $1.79 | 3k | 2026-09-03 |
| 9 | Muse Wondermax | 30.6% | 61.9 | $0.723 | 11k | 2026-10-12 |
| 10 | GPT-6 Astramedium | 30.4% | 64.5 | $1.72 | 2k | 2026-09-03 |
| 11 | GPT-6 Astralow | 30.4% | 62.2 | $1.70 | 1k | 2026-09-03 |
| 12 | GPT-6.1 Solmedium | 30.0% | 62.9 | $0.337 | 1k | 2026-09-29 |
| 13 |  Gemini 4 Carbonmedium | 29.6% | 64.1 | $0.480 | 7k | 2026-10-28 |
| 14 | Muse Wonderhigh | 29.2% | 56.3 | $0.599 | 5k | 2026-10-12 |
| 15 |  Claude Opus 5.5high | 28.8% | 65.4 | $0.825 | 6k | 2026-09-22 |
| 16 | GPT-6.1 Terramax | 28.6% | 65.0 | $0.235 | 15k | 2026-10-26 |
| 17 | Gemini 4 Carbonlow | 28.6% | – | $0.381 | 2k | 2026-10-28 |
| 18 | Gemini 4 Carbonhigh | 28.4% | 68.5 | $0.847 | 23k | 2026-10-28 |
| 19 | GPT-6.1 Terramedium | 28.4% | 58.6 | $0.167 | 1k | 2026-10-26 |
| 20 | GPT-6.1 Terrahigh | 28.0% | 59.7 | $0.171 | 2k | 2026-10-26 |
| 21 | Claude Fable 5.1low | 28.0% | 58.7 | $1.91 | 4k | 2026-09-01 |
| 22 | GPT-5.6 Solhigh | 27.8% | 63.5 | – | – | 2026-07-09 |
| 23 | GPT-5.6 Solxhigh | 27.6% | 64.2 | – | – | 2026-07-09 |
| 24 | GPT-5.6 Solmax | 27.2% | 65.4 | – | – | 2026-07-09 |
| 25 | GPT-6.1 Sollow | 27.0% | 60.2 | $0.334 | 1k | 2026-09-29 |
| 26 | NewClaude Fable 5.2low | 26.8% | 65.3 | $1.53 | 3k | 2026-11-09 |
| 27 | Claude Fable 5.1high | 26.8% | 63.3 | $2.00 | 6k | 2026-09-01 |
| 28 | Claude Fable 5.1medium | 26.8% | 61.7 | $1.95 | 4k | 2026-09-01 |
| 29 | Claude Opus 5.5xhigh | 26.6% | 65.5 | $0.964 | 13k | 2026-09-22 |
| 30 | Muse Spark 1.3max | 26.6% | 61.7 | $0.205 | 11k | 2026-09-02 |
| 31 | GPT-6.1 Terraxhigh | 26.6% | 60.1 | $0.182 | 4k | 2026-10-26 |
| 32 | Claude Opus 5.5max | 26.2% | 68.9 | $1.55 | 42k | 2026-09-22 |
| 33 | Claude Fable 5.1max | 26.2% | 66.9 | $2.76 | 20k | 2026-09-01 |
| 34 | Claude Fable 5.1xhigh | 26.2% | 64.6 | $2.31 | 12k | 2026-09-01 |
| 35 | GPT-5.6 Solmedium | 26.2% | 61.4 | – | – | 2026-07-09 |
| 36 | Claude Sonnet 5.5max | 25.8% | 65.0 | $0.794 | 44k | 2026-09-28 |
| 37 | Claude Opus 5.5medium | 25.6% | 61.5 | $0.795 | 4k | 2026-09-22 |
| 38 | Claude Opus 5.5low | 25.6% | 59.0 | $0.764 | 3k | 2026-09-22 |
| 39 | GPT-6.1 Terralow | 25.6% | 56.7 | $0.164 | 612 | 2026-10-26 |
| 40 | GPT-6 Solmax | 25.2% | 62.9 | $0.425 | 10k | 2026-09-22 |
| 41 | New Inkling 2 Small | 25.2% | 58.3 | $0.057 | 24k | 2026-11-05 |
| 42 | Claude Sonnet 5.5high | 25.2% | 58.0 | $0.409 | 6k | 2026-09-28 |
| 43 | Claude Sonnet 5.5xhigh | 24.6% | 62.7 | $0.460 | 11k | 2026-09-28 |
| 44 | GPT-5.6 Terraxhigh | 24.6% | 60.3 | – | – | 2026-07-09 |
| 45 | GPT-6 Solxhigh | 24.6% | 58.6 | – | – | 2026-09-22 |
| 46 | GPT-6 Solhigh | 24.4% | 58.2 | – | – | 2026-09-22 |
| 47 | NewClaude Fable 5.2max | 24.2% | 69.4 | $2.52 | 28k | 2026-11-09 |
| 48 | Muse Spark 1.3xhigh | 24.2% | 58.1 | – | – | 2026-09-02 |
| 49 | GPT-6 Sollow | 24.2% | 54.6 | – | – | 2026-09-22 |
| 50 | Claude Fable 5max | 24.0% | 65.3 | – | – | 2026-06-09 |
| 51 | GPT-5.6 Terramax | 24.0% | 60.5 | $0.589 | 22k | 2026-07-09 |
| 52 | GPT-5.6 Lunamax | 24.0% | 56.2 | $0.057 | 20k | 2026-07-09 |
| 53 | NewClaude Fable 5.2medium | 23.8% | 66.3 | $1.56 | 4k | 2026-11-09 |
| 54 | GPT-5.6 Lunaxhigh | 23.8% | 58.4 | – | – | 2026-07-09 |
| 55 | GPT-6 Solmedium | 23.8% | 57.0 | – | – | 2026-09-22 |
| 56 | Gemini 3.7 Flashhigh | 23.6% | 62.2 | – | – | 2026-08-13 |
| 57 |  Grok 4.7high | 23.2% | 57.0 | $0.302 | 13k | 2026-09-21 |
| 58 | GPT-6.1 Terrano reasoning | 23.2% | 51.9 | $0.164 | 629 | 2026-10-26 |
| 59 | Gemini 3.8 Flashmedium | 22.8% | 60.3 | $0.087 | 6k | 2026-09-02 |
| 60 | Claude Opus 4.8max | 22.8% | 60.2 | – | – | 2026-05-28 |
More agents & tools benchmarks · All LLM benchmarks
Related benchmarks
Top head-to-heads on GDP.pdf (AA)
Cite as: BenchLeader, “GDP.pdf (AA) leaderboard”, data as of 9 Nov 2026, https://www.benchleader.com/benchmarks/aa_gdp_pdf.
GDP.pdf (AA): questions
- What does GDP.pdf (AA) measure?
- Document work on real PDFs, scored all-or-nothing: every required element has to be right for the task to pass, which is why the scores look low. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on GDP.pdf (AA)?
- As of 9 Nov 2026, the highest score on GDP.pdf (AA) is 32.6% by Muse Wonder Full (result dated 12 Oct 2026), ahead of GPT-6 Astra at 32.2%. Scores come from Artificial Analysis and are re-read daily.
- Which AI model leads GDP.pdf (AA)?
- Muse Wonder Full leads GDP.pdf (AA) with 32.6% as of 9 Nov 2026 (result dated 12 Oct 2026), ahead of GPT-6 Astra at 32.2%.
- How many models have GDP.pdf (AA) results?
- 252 model configurations have a GDP.pdf (AA) result on BenchLeader, all taken from Artificial Analysis.
- Who runs GDP.pdf (AA) and how often is it updated?
- GDP.pdf (AA) is published by Artificial Analysis. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does GDP.pdf (AA) count toward the BenchLeader Index?
- No. GDP.pdf (AA) is shown for reference but left out of the composite index.