Gemini 3.1 Pro Preview vs Kimi K2 Thinking
Gemini 3.1 Pro Preview edges ahead on overall intelligence (RunFree 78 vs 45.2). Here's how they stack up on benchmarks, price and specs.
Gemini 3.1 Pro Preview
Google
Kimi K2 Thinking
MoonshotAI
Shared benchmarks (5)
Loading chart…
| Gemini 3.1 Pro Preview | Kimi K2 Thinking | |
|---|---|---|
| RunFree Score | 78 | 45.2 |
| Blended price / 1M | $4.50 | $1.07 |
| Input / 1M | $2.00 | $0.60 |
| Output / 1M | $12.00 | $2.50 |
| Context window | 1.0M | 262K |
| Max output | 66K | 262K |
| Reasoning model | Yes | Yes |
| GPQA Diamond | 94.3% | 84.5% |
| Humanity's Last Exam | 44.4% | 23.9% |
| MMLU-Pro | 92.6% | 84.6% |
| SWE-Bench Verified | 80.6% | 71.3% |
| Terminal-Bench | 68.5% | 47.1% |
| Wins | 7 | 4 |
A blank means the metric is tied or one model has no verified data. Prices are blended 3:1 input:output.
Compare more: Gemini 3.1 Pro Preview card · Kimi K2 Thinking card · full leaderboard