Gemini 3 Flash vs moonshotai/Kimi-K2.5
moonshotai/Kimi-K2.5 is cheaper on output tokens, while Gemini 3 Flash offers a larger context window. Choose Gemini 3 Flash or moonshotai/Kimi-K2.5 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Gemini 3 Flash | moonshotai/Kimi-K2.5 |
|---|---|---|
| Provider | OpenCode | SiliconFlow |
| Input / 1M tokens | $0.50 | $0.45 |
| Output / 1M tokens | $3.00 | $2.25 |
| Context window | 1M | 262K |
| Parameters | — | 1T |
| Open weights | No | Yes |
| Released | Dec 2025 | Jan 2026 |
| SWE-bench Multilingual score | 72.7 | — |
| LMArena (Chatbot Arena) Elo score | 1466 | — |
| LMArena WebDev Elo score | 1440 | — |
| LMArena Vision Elo score | 1285 | — |
| Epoch Capabilities Index score | 152 | — |
FAQ
Is Gemini 3 Flash or moonshotai/Kimi-K2.5 cheaper?
moonshotai/Kimi-K2.5 is cheaper on output tokens ($2.25 vs $3.00 per 1M).
Which has the larger context window, Gemini 3 Flash or moonshotai/Kimi-K2.5?
Gemini 3 Flash has the larger context window (1M tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.