Multimodal · Benchmark profile
Tool-free variant of CharXiv that isolates raw visual reasoning ability without code execution or tool augmentation.
Data verified 21 Jul 2026 · Methodology 1.6.0
Benchmark score on CharXiv Reasoning without tools
Claude Mythos 5 leads at 88.9%, followed by Kimi K3 (84.8%) and Claude Opus 4.7 (Adaptive) (82.1%).
Visual analysis
Switch between model placement, score distribution and descriptive provider averages. Every view uses the same sourced leaderboard.
6 ranked models · higher is better
| Model | Rank | Provider | Score |
|---|---|---|---|
| Claude Mythos 5 | #1 | Anthropic | 88.9% |
| Kimi K3 | #2 | Moonshot AI | 84.8% |
| Claude Opus 4.7 (Adaptive) | #3 | Anthropic | 82.1% |
| Claude Opus 4.8 | #4 | Anthropic | 80.5% |
| Inkling | #5 | Thinking Machines Lab | 78.1% |
| Claude Sonnet 5 | #6 | Anthropic | 77% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | Claude Mythos 5 claude-mythos-5 | Anthropic | closed | Reference only | 88.9% |
| #2 | Kimi K3 kimi-k3 | Moonshot AI | closed | Reference only | 84.8% |
| #3 | Claude Opus 4.7 (Adaptive) claude-opus-4-7-max | Anthropic | closed | Estimated reference | 82.1% |
| #4 | Claude Opus 4.8 claude-opus-4-8 | Anthropic | closed | Reference only | 80.5% |
| #5 | Inkling inkling | Thinking Machines Lab | open | Reference only | 78.1% |
| #6 | Claude Sonnet 5 claude-sonnet-5 | Anthropic | closed | Estimated reference | 77% |
The top of this snapshot is led by Claude Mythos 5 at 88.9%; third place is 6.8 points behind. The top-6 spread is 11.9 points.
About CharXiv Reasoning without tools
Tool-free variant of CharXiv that isolates raw visual reasoning ability without code execution or tool augmentation. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
Tool-free variant of CharXiv that isolates raw visual reasoning ability without code execution or tool augmentation.
Claude Mythos 5 by Anthropic currently leads with 88.9%.
6 models in the LuminaBench cohort have a qualifying score on this benchmark.
No. This benchmark is display-only and does not enter the overall Lumina composite.
Related
