Tools
Compare models
Add up to four models via ?models=slug-a,slug-b — or pick from any model page.
| Spec | Moonshot AI |
|---|---|
| Params | 1.1T / 32B |
| Architecture | moe |
| Context | 262K |
| License | |
| License terms | Custom licence |
| 667.5 GBest | |
| Cheapest hosted (out/1M) | $1.83 · OpenRouter |
| Released | Apr 14, 2026 |
| LiveBench | 70.54 |
| LiveBench Agentic Coding | 46.92 |
| LiveBench Coding | 78.57 |
| LiveBench Data Analysis | 65.13 |
| LiveBench Instruction Following | 64.36 |
| LiveBench Language | 75.14 |
| LiveBench Mathematics | 84.28 |
| LiveBench Reasoning | 79.38 |
| Arena Agent | 28th of 55 |
| Arena Agent · Recoverygets back on track after a command fails | 46th of 55 |
| Arena Agent · Steerabilitydoes what it was asked, and changes course when told | 5th of 55 |
| Arena Agent · Task outcomefinishes what the session set out to do | 38th of 55 |
| Arena Agent · Tool usereaches for the right one, and does not invent one | 2nd of 55 |
| Arena Coding | 1516 |
| Arena Creative Writing | 1433 |
| Arena Hard Prompts | 1485 |
| Arena Instruction Following | 1456 |
| Arena Maths | 1480 |
| Arena Text (overall) | 1461 |
| Arena Code (WebDev) | 1509 |
