Kimi K2 Thinking vs DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 is cheaper on output tokens, while DeepSeek V4 Flash 0731 offers a larger context window. Choose Kimi K2 Thinking or DeepSeek V4 Flash 0731 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Kimi K2 Thinking | DeepSeek V4 Flash 0731 |
|---|---|---|
| Provider | Charm Hyper | Charm Hyper |
| Input / 1M tokens | $0.60 | $0.44 |
| Output / 1M tokens | $2.50 | $1.32 |
| Context window | 262K | 1M |
| Parameters | — | — |
| Open weights | Yes | Yes |
| Released | Sep 2026 | Aug 2026 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.