Anthropic
availableShows if the model has enough results for an index.Claude Mythos 5
Claude Mythos 5 is a reasoning model from Anthropic in the Claude Mythos family. 8 benchmarks count toward its score, in 4 categories.
IndexOverall score. 50 is the middle.72.0 ±9.2
CoverageShare of the index weight with results.65%
SpeedOutput tokens per second.—
Input / 1MUS dollars per 1M input tokens.$10
Output / 1MUS dollars per 1M output tokens.$50
ContextMaximum tokens in one request.1M
EloLMArena rating and rank.N/A
50 is the middle of the board. The range shows the doubt in the index.
CapabilitiesScore per category. 50 is the middle.
50 is the middleResults
8 counted| BenchmarkThe test name. | CategoryThe capability that the test measures. | ResultThe score from the publisher. | IndexThis result on the index scale. | RunThe settings of the run. | DateDate of the result. | Published byThe source of the result. |
|---|---|---|---|---|---|---|
| United States of America Mathematical Olympiad 2026 | Math | 97.6% | — | — | — | Mathematical Association of America |
| Software Engineering Benchmark Verified | Coding | 95.5% | 74.2 | — | — | Carlos E. Jimenez et al. |
| Graduate-Level Google-Proof Q&A | Knowledge | 94.1% | 65.1 | — | — | David Rein et al. |
| CharXiv Reasoning | Multimodal | 93.5% | 72.6 | — | — | CharXiv authors |
| CharXiv Reasoning without tools | Multimodal | 88.9% | — | — | — | CharXiv authors |
| BrowseComp | Agentic | 88.0% | 73.6 | — | — | OpenAI |
| OSWorld-Verified | Agentic | 85.0% | 72.9 | — | — | Tianbao Xie et al. |
| CyberGym | Agentic | 83.8% | 73.5 | — | — | Zhun Wang et al. |
| SWE-bench Pro | Coding | 80.3% | 81.7 | — | — | Xiang Deng et al. |
| Humanity's Last Exam | Knowledge | 64.5% | 83.4 | — | — | Center for AI Safety et al. |
| Humanity's Last Exam without tools | Knowledge | 59.0% | 78.8 | — | — | OpenAI |
8 benchmarks count, from 9 of 11 results. A grey row does not count. Too few models took that benchmark.