Chartography leaderboard
As of November 8, 2026, Kimi K3.1 Pro (Moonshot) leads Chartography at 90.2%, ahead of Claude Opus 5.5 at 89% and Claude Fable 5.1 at 88.4%. 10 systems ranked. Every score here is published by the model's own lab; each row links to the page it comes from.
What is Chartography? Chartography tests whether a model can read the charts professionals use, in medicine, engineering, finance, manufacturing and science, and answer questions that need several readings and arithmetic. Made by Surge AI, August 2026.
Top models on Chartography, November 2026
The model may run code or image tools.
| # | Model | Score | Price / 1M | Context | Source |
|---|---|---|---|---|---|
| 1 | Kimi K3.1 Pro Moonshot (Kimi) | max effort, with Python | $2 / $10 | 1M | Vendor-reported Moonshot AI |
| 2 | Claude Opus 5.5 Anthropic | with tools | $4 / $20 | 1M | Vendor-reported Anthropic |
| 3 | Claude Fable 5.1 Anthropic | with tools | $10 / $50 | 1M | Vendor-reported Anthropic |
| 4 | Claude Opus 5 Anthropic | with tools | $5 / $25 | 1M | Vendor-reported Anthropic |
| 5 | DeepSeek V4.1 Flash DeepSeek · listed as “DeepSeek V4.1 Flash (1017)” | with tools | $0.3 / $1.2 | 1M | Vendor-reported DeepSeek |
Chartography under other settings
Labs report Chartography under different settings, and a score from one setting is not comparable with a score from another. Each setting has its own table, ranked on its own.
No tools
| # | Model | Score | Price / 1M | Context | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 Anthropic | no tools | $4 / $20 | 1M | Vendor-reported Anthropic |
| 2 | Claude Sonnet 5.5 Anthropic | no tools | $2 / $10 | 1M | Vendor-reported Anthropic |
| 3 | Claude Haiku 5.5 Anthropic | no tools | $0.1 / $0.5 | 1M | Vendor-reported Anthropic |
| 4 | Claude Sonnet 5 Anthropic | no tools | $2 / $10 | 1M | Vendor-reported Anthropic |
| 5 | Claude Haiku 4.5 Anthropic | no tools | $1 / $5 | 200K | Vendor-reported Anthropic |
Tool use not stated
| # | Model | Score | Price / 1M | Context | Source |
|---|---|---|---|---|---|
| 1 | Gemini 4 Argon Google DeepMind | tool use not stated | $2 / $10 | 1M | Vendor-reported Google DeepMind |
Chartography: score against price
Can you trust Chartography scores?
No independent leaderboard publishes Chartography results for these models yet, so every number on this page is the lab's own report of its own model. Labs choose the harness, the effort level and the subset they report, and they rarely publish failures. Read the setting under each score before comparing two rows.
- In the authors' own evaluation of 30 model configurations the best reached 45.0% mean pass@1. Labs report much higher numbers from their own harnesses and tool settings, so lab scores and the paper's scores are not the same measurement.
- The tasks are public, so later models may have seen them.
- Surge AI supplies training data to AI labs and also built and publishes this benchmark.
How Chartography works
The task
Each task shows a domain-specific chart and asks a question written by someone who reads such charts for a living; the answer is a number, a range or a set of labels.
The score
Answers are checked against an expert answer with an acceptable range set by experts for each task. The authors report mean pass@1 over 20 scored trials.
Test set
100 tasks, as published by the authors.
Made by
Surge AI, first published August 2026.
Official sources
- surgehq.ai/blog/chartography
- Paper: arxiv.org/abs/2608.10677
- Definition read on 2026-11-08
Sources and freshness
12 published scores from 7 sources. Each was read on its source page; the oldest reading on this page is from November 8, 2026. Prices beside each score come from our model price index and move when a vendor reprices.
- Vendor-reported Anthropic
- Vendor-reported Anthropic
- Vendor-reported Anthropic
- Vendor-reported DeepSeek
- Vendor-reported DeepSeek
- Vendor-reported Google DeepMind
- Vendor-reported Moonshot AI
Chartography FAQ
Answers derived from the 12 published scores on this page and the current price index.