Tools
Compare models
Add up to four models via ?models=slug-a,slug-b — or pick from any model page.
| Spec | Moonshot AI | Moonshot AI |
|---|---|---|
| Params | 1.1T / 32B | 1.1T / 32B |
| Architecture | moe | moe |
| Context | 262K | 262K |
| Weights | ||
| License | Custom licence | Custom licence |
| 667.5 GBest | 667.5 GBest | |
| Cheapest hosted (out/1M) | $2.48 · OpenRouter | $3.50 · OpenRouter |
| Released | Apr 14, 2026 | Jun 11, 2026 |
| LiveBench | 70.5via Thinking | 68.4 |
| LiveBench Agentic Coding | 46.9via Thinking | 45.7 |
| LiveBench Coding | 78.6via Thinking | 74 |
| LiveBench Data Analysis | 65.1via Thinking | 62.7 |
| LiveBench Instruction Following | 64.4via Thinking | 56.3 |
| LiveBench Language | 75.1via Thinking | 77.9 |
| LiveBench Mathematics | 84.3via Thinking | 79.6 |
| LiveBench Reasoning | 79.4via Thinking | 82.8 |
| Arena Agent (IPS) | −0.011 | 0.004 |
| Arena Coding | 1514.8 | — |
| Arena Creative Writing | 1429.5 | — |
| Arena Hard Prompts | 1484.8 | — |
| Arena Instruction Following | 1454.2 | — |
| Arena Maths | 1479.3 | — |
| Arena Text (overall) | 1460.7 | — |
| Arena Code (WebDev) | 1509.5 | 1472.6 |
