Agentic benchmark
GDP.pdf leaderboard
Artificial Analysis GDP.pdf. Every model the catalog carries a published GDP.pdf value for, ranked by that value.
Artificial Analysis' document-output benchmark: professional tasks whose deliverable is a produced PDF, one of the ten components of its Intelligence Index v4.3.
GDP.pdf ranking
14 models with a published GDP.pdf value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Task success rate |
|---|---|---|---|
| 1 |  GPT-6 Astra | OpenAI | 31.0 |
| 2 | GPT-5.6 Sol | OpenAI | 27.2 |
| 3 |  Muse Spark 1.3 | Meta | 26.6 |
| 4 |  Claude Fable 5.1 | Anthropic | 26.2 |
| 5 | Claude Fable 5 | Anthropic | 24.0 |
| 5 | GPT-5.6 Luna | OpenAI | 24.0 |
| 5 | GPT-5.6 Terra | OpenAI | 24.0 |
| 8 |  Kimi K3 | Moonshot AI | 22.0 |
| 9 | Claude Opus 5 | Anthropic | 21.6 |
| 10 |  Gemini 3.8 Flash | 21.0 | |
| 11 |  Grok 4.6 | xAI | 17.0 |
| 12 |  Qwen3.8-27B | Alibaba | 16.4 |
| 13 |  GLM-5.3-Flash | Z.AI | 15.4 |
| 14 | Gemini 3.5 Flash-Lite | 13.6 |
Evidence key: Observed
Rows are ordered by the value GDP.pdf Benchmark Leaderboard published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.