Kimi K2.7 Code vs Gemini 3 Flash Thinking
Gemini 3 Flash Thinking is cheaper on output tokens, while Gemini 3 Flash Thinking offers a larger context window. Choose Kimi K2.7 Code or Gemini 3 Flash Thinking based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Kimi K2.7 Code | Gemini 3 Flash Thinking |
|---|---|---|
| Provider | MoonshotAI | NanoGPT |
| Input / 1M tokens | $0.56 | $0.50 |
| Output / 1M tokens | $3.10 | $3.00 |
| Context window | 262K | 1M |
| Parameters | — | — |
| Open weights | Yes | No |
| Released | Jun 2026 | Dec 2025 |
| GPQA Diamond score | 87.9 | — |
| Epoch Capabilities Index score | 150 | — |
| LMArena WebDev Elo score | 1474 | — |
| LiveBench score | 68.4 | 73.0 |
FAQ
Is Kimi K2.7 Code or Gemini 3 Flash Thinking cheaper?
Gemini 3 Flash Thinking is cheaper on output tokens ($3.00 vs $3.10 per 1M).
Which has the larger context window, Kimi K2.7 Code or Gemini 3 Flash Thinking?
Gemini 3 Flash Thinking 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.