Multimodal & Grounded benchmark
CharXiv w/o tools leaderboard
CharXiv Reasoning without tools. Every model the catalog carries a published CharXiv w/o tools value for, ranked by that value.
Tool-free variant of CharXiv that isolates raw visual reasoning ability without code execution or tool augmentation.
CharXiv w/o tools ranking
14 models with a published CharXiv w/o tools 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 | Chart understanding without tools |
|---|---|---|---|
| 1 |  Claude Mythos 5 | Anthropic | 88.9 |
| 2 |  Qwen3.8 Max | Alibaba | 88.4 |
| 3 |  Gemini 3.8 Flash | 86.2 | |
| 4 |  Kimi K3 | Moonshot AI | 84.8 |
| 5 | Qwen3.8-Flash-Next | Alibaba | 84.6 |
| 6 | Gemini 3.7 Flash | 84.5 | |
| 7 | Qwen3.8-27B | Alibaba | 83.7 |
| 8 | Qwen3.8-Omni-Flash | Alibaba | 83.5 |
| 9 | DSdots3-note Preview | Dots Studio | 83.1 |
| 10 | Claude Opus 4.7 (Adaptive) | Anthropic | 82.1 |
| 11 | Claude Opus 4.8 | Anthropic | 80.5 |
| 12 | TMInkling | Thinking Machines Lab | 78.1 |
| 13 | TMInkling-Small | Thinking Machines Lab | 77.4 |
| 14 | Claude Sonnet 5 | Anthropic | 77 |
Evidence key: Observed
Rows are ordered by the value CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMs 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.