3 vs 0 benchmarks won
Google Gemini 3.0 Pro | Moonshot AI Kimi K2 Thinking | |
|---|---|---|
| Overview | ||
| Company | Moonshot AI | |
| Release date | Nov 18 2025 | Nov 6 2025 |
| Model type | — | — |
| Open source | No | Yes |
| Specifications | ||
Parameters | — | 1T |
Context window | — | 256k |
| Benchmarks | ||
Science reasoning GPQA Diamond | 91.9% | — |
Software engineering SWE-Bench Verified | 76.2% | 71.3% |
Multimodal understanding MMMU | 81% | — |
| Timeline | ||
| Release gap | Kimi K2 Thinking shipped 12 days before Gemini 3.0 Pro | |
Gemini 3.0 Pro leads Kimi K2 Thinking on 3 of the tracked benchmarks (GPQA Diamond, SWE-Bench Verified, MMMU). Kimi K2 Thinking shipped 12 days before Gemini 3.0 Pro, so benchmark comparisons should account for the intervening progress.
Kimi K2 Thinking is an open-source / open-weight model; Gemini 3.0 Pro is proprietary.
On SWE-Bench Verified, Gemini 3.0 Pro scores 76.2%, 4.9 points above Kimi K2 Thinking at 71.3%.