Kimi K2.5 vs DeepSeek V4 Pro 0813
Kimi K2.5 is cheaper on output tokens, while DeepSeek V4 Pro 0813 offers a larger context window. Choose Kimi K2.5 or DeepSeek V4 Pro 0813 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Kimi K2.5 | DeepSeek V4 Pro 0813 |
|---|---|---|
| Provider | TensorX | TensorX |
| Input / 1M tokens | $0.50 | $2.00 |
| Output / 1M tokens | $2.80 | $4.00 |
| Context window | 262K | 1M |
| Parameters | — | — |
| Open weights | Yes | Yes |
| Released | Jan 2026 | Aug 2026 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.