Reported by others
Reading PDFs (GDP.pdf)
The model reads real-world PDF documents with tables and figures and answers questions about their contents.
How to read this table
- Higher rows are rated higher. A range like “1–2” means the gap is within the margin of error.
- The score is the value published by the task's authors and other organizations.
- Conditions can differ between models, so small gaps do not mean much.
- Prices are a guide per million tokens, reading then writing.
- Length is how much text the model can read at once (tokens).
| Rank | Model | Developer | Accuracy (%) | Input $ /1M | Output $ /1M | Context | Supports | Weights | Released | Price and specs |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | GPT-6.1 Sol | OpenAI | Rating 35.0 | $2 | $10 | 1.05M | OpenAI Released 2026-09 Price $2 / $10 Length 1.05M | |||
| 2 | GPT-6 Astra | OpenAI | Rating 34.2 | $10 | $50 | 1.05M | OpenAI Released 2026-09 Price $10 / $50 Length 1.05M | |||
| 3 | Claude Sonnet 5.5 | Anthropic | Rating 31.4 | $2 | $10 | 1M | Anthropic Released 2026-09 Price $2 / $10 Length 1M | |||
| 4 | Muse Wonder | Meta | Rating 30.8 | $4 | $20 | 1.05M | Meta Released 2026-10 Price $4 / $20 Length 1.05M | |||
| 5 | GPT-5.6 Sol | OpenAI | Rating 30.7 | $4 | $20 | 1.05M | OpenAI Released 2026-07 Price $4 / $20 Length 1.05M | |||
| 6 | Claude Opus 5.5 | Anthropic | Rating 30.6 | $4 | $20 | 1M | Anthropic Released 2026-09 Price $4 / $20 Length 1M | |||
| 7 | Claude Fable 5 | Anthropic | Rating 30.0 | $10 | $50 | 1M | Anthropic Released 2026-06 Price $10 / $50 Length 1M | |||
| 8 | Claude Fable 5.1 | Anthropic | Rating 29.6 | $10 | $50 | 1M | Anthropic Released 2026-09 Price $10 / $50 Length 1M | |||
| 9 | Gemini 4 Argon | Rating 28.4 | $2 | $10 | 1.05M | Google Released 2026-09 Price $2 / $10 Length 1.05M | ||||
| 10 | Muse Spark 1.3 | Meta | Rating 27.6 | $1.25 | $4.25 | 1.05M | Meta Released 2026-09 Price $1.25 / $4.25 Length 1.05M | |||
| 11 | GPT-6 Sol | OpenAI | Rating 26.4 | $2 | $10 | 1.05M | OpenAI Released 2026-09 Price $2 / $10 Length 1.05M | |||
| 12 | GPT-5.5 | OpenAI | Rating 26.0 | $5 | $30 | 1.05M | OpenAI Released 2026-04 Price $5 / $30 Length 1.05M | |||
| 13 | GPT-5.6 Terra | OpenAI | Rating 24.7 | $2 | $12 | 1.05M | OpenAI Released 2026-07 Price $2 / $12 Length 1.05M | |||
| 14–15 | Claude Opus 4.8 | Anthropic | Rating 24.0 | $5 | $25 | 1M | Anthropic Released 2026-05 Price $5 / $25 Length 1M | |||
| 14–15 | Claude Opus 5 | Anthropic | Rating 24.0 | $5 | $25 | 1M | Anthropic Released 2026-07 Price $5 / $25 Length 1M | |||
| 16 | Gemini 3.7 Flash | Rating 23.8 | $0.75 | $3.75 | 1.05M | Google Released 2026-08 Price $0.75 / $3.75 Length 1.05M | ||||
| 17 | Gemini 3.8 Flash | Rating 23.4 | $0.75 | $3.75 | 1.05M | Google Released 2026-09 Price $0.75 / $3.75 Length 1.05M | ||||
| 18 | Qwen3.8 Max | Alibaba | Rating 23.2 | $2 | $6 | 1M | Alibaba Released 2026-08 Price $2 / $6 Length 1M | |||
| 19 | GPT-6 Luna | OpenAI | Rating 23.0 | $0.1 | $0.5 | 1.05M | OpenAI Released 2026-09 Price $0.1 / $0.5 Length 1.05M | |||
| 20 | Grok 4.7 | xAI | Rating 22.8 | $2 | $6 | 500K | xAI Released 2026-09 Price $2 / $6 Length 500K | |||
| 21 | GPT-5.6 Luna | OpenAI | Rating 22.7 | $0.2 | $1.2 | 1.05M | OpenAI Released 2026-07 Price $0.2 / $1.2 Length 1.05M | |||
| 22 | Claude Opus 4.7 | Anthropic | Rating 21.0 | $5 | $25 | 1M | Anthropic Released 2026-04 Price $5 / $25 Length 1M | |||
| 23 | DeepSeek-V4.1 Flash | DeepSeek | Rating 19.8 | — | — | 1M | DeepSeek Released 2026-09 Price — Length 1M | |||
| 24 | Kimi K3 | Moonshot | Rating 19.0 | $3 | $15 | 1.05M | Moonshot Released 2026-07 Price $3 / $15 Length 1.05M | |||
| 25 | Claude Sonnet 4.6 | Anthropic | Rating 18.0 | $3 | $15 | 1M | Anthropic Released 2026-02 Price $3 / $15 Length 1M | |||
| 26 | Grok 4.6 | xAI | Rating 17.2 | $2 | $6 | 500K | xAI Released 2026-08 Price $2 / $6 Length 500K | |||
| 27 | Gemini 3.1 Pro (preview) | Rating 17.0 | $2 | $12 | 1.05M | Google Released 2026-02 Price $2 / $12 Length 1.05M | ||||
| 28 | Qwen3.8 Flash | Alibaba | Rating 16.6 | $0.15 | $0.47 | 1M | Alibaba Released 2026-08 Price $0.15 / $0.47 Length 1M | |||
| 29 | Muse Spark 1.2 | Meta | Rating 16.0 | $1.25 | $4.25 | 1.05M | Meta Released 2026-08 Price $1.25 / $4.25 Length 1.05M | |||
| 30 | Muse Spark 1.1 | Meta | Rating 15.0 | $1.25 | $4.25 | 1M | Meta Released 2026-07 Price $1.25 / $4.25 Length 1M | |||
| 31–34 | Gemini 3.5 Flash | Rating 14.0 | $1.5 | $9 | 1.05M | Google Released 2026-05 Price $1.5 / $9 Length 1.05M | ||||
| 31–34 | Gemini 3.6 Flash | Rating 14.0 | $0.75 | $3.75 | 1.05M | Google Released 2026-07 Price $0.75 / $3.75 Length 1.05M | ||||
| 31–34 | GLM-5.3 Flash | Z.ai | Rating 14.0 | $0.15 | $0.5 | 1M | Z.ai Released 2026-08 Price $0.15 / $0.5 Length 1M | |||
| 31–34 | Grok 4.5 | xAI | Rating 14.0 | $2 | $6 | 500K | xAI Released 2026-07 Price $2 / $6 Length 500K | |||
| 35 | Kimi K2.6 | Moonshot | Rating 12.0 | $0.95 | $4 | 262.1K | Moonshot Released 2026-04 Price $0.95 / $4 Length 262.1K | |||
| 36–37 | Gemini 3.5 Flash-Lite | Rating 10.0 | $0.3 | $2.5 | 1.05M | Google Released 2026-07 Price $0.3 / $2.5 Length 1.05M | ||||
| 36–37 | Gemini 3 Flash (preview) | Rating 10.0 | $0.5 | $3 | 1M | Google Released 2025-12 Price $0.5 / $3 Length 1M | ||||
| 38 | Grok 4.3 | xAI | Rating 8.0 | $1.25 | $2.5 | 1M | xAI Released 2026-04 Price $1.25 / $2.5 Length 1M | |||
| 39 | Amazon Nova 2 Pro (preview) | Amazon | Rating 2.0 | — | — | — | Amazon Released 2025-12 Price — Length — |
Best accuracy over time
2025-12-02Release date2026-10-12
How to read this page
- The value is the share answered correctly. Figures published by the organizations that made the tasks, among others, collected here.
- The accuracy is the share of problems solved. A gap of 1 point is 1 question out of 100.
- In the Supports column the icons are, from left: can use tools / can read images / can answer in a fixed format (JSON); faded means not supported or unknown.
- Open-weights models have their contents (the weights) published, so you can also run them on your own computer or server. Models without the open-weights mark are normally used through the developer's own services (apps or API); for some models the weights status is unknown.
- Names such as ChatGPT next to a model are the consumer services its developer offers. Check on the service itself whether that model is available there.
Data: Arena, Epoch AI, Swallow LLM team (CC BY 4.0), Models.dev (MIT), updated 2026-10-24, data sources