In short
How they differ.
Worked out from the figures below. Prices and specs are the makers' own, read Oct 20, 2026.
- Kimi K3 can be downloaded and run on your own machine; Claude Mythos 5.1 can't.
- No independent board we follow ranks both models yet.
Side by side
The specs.
| Claude Mythos 5.1 | Kimi K3 | |
|---|---|---|
| Made by | Anthropic · US | Moonshot AI · China |
| Released | Sep 1, 2026 | Not stated |
| Price per 1M tokens | No public price | $3 in / $15 out |
| Context window | Not stated | 1M |
| Takes in | Not stated | text, image, video |
| Download and run it yourself | No | Yes (Kimi K3 License) |
| Size | Not published | 2.8T parameters, 104B active |
| Maker says | “Our most capable model for cybersecurity and biology research.” | “Kimi K3 is an open-weight, native multimodal agentic model and our most capable model to date.” |
| Sources | Prices, Oct 20, 2026 · Description | Prices, Oct 20, 2026 · Description |
Independent tests
Where they place.
Position on each board we follow (1 is best). Epoch AI's and Vectara's scores are shown; the other boards are linked for their own figures.
| Board | Claude Mythos 5.1 | Kimi K3 |
|---|---|---|
| Epoch Capabilities IndexEpoch AI, Oct 20, 2026 | Not listed | #17157.42 |
| LMArena textLMArena, Oct 15, 2026 | Not listed | #18 |
| LMArena coding chatsLMArena, Oct 15, 2026 | Not listed | #12 |
| LMArena WebDevLMArena, Oct 19, 2026 | Not listed | #16 |
| LMArena mathLMArena, Oct 15, 2026 | Not listed | #16 |
| Humanity's Last ExamScale AI | Not listed | #14 |
| SWE-Bench ProScale AI, Sep 22, 2026 | Not listed | #4 |
| Vals IndexVals AI, Oct 19, 2026 | Not listed | #18 |
| Vibe Code BenchVals AI, Oct 18, 2026 | Not listed | #16 |
| ProofBenchVals AI, Oct 12, 2026 | Not listed | #12 |
| GPQA Diamond (run by Vals AI, archived)Vals AI, Sep 1, 2026 | Not listed | #12 |
| MMMU Pro (run by Vals AI, archived)Vals AI, Sep 1, 2026 | Not listed | #11 |
More comparisons