DeepSeek V4 Pro vs DeepSeek V3.2
DeepSeek V3.2 is cheaper on output tokens, while DeepSeek V4 Pro offers a larger context window. Choose DeepSeek V4 Pro or DeepSeek V3.2 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | DeepSeek V4 Pro | DeepSeek V3.2 |
|---|---|---|
| Provider | TensorX | TensorX |
| Input / 1M tokens | $1.75 | $0.30 |
| Output / 1M tokens | $3.50 | $0.50 |
| Context window | 1M | 164K |
| Parameters | — | — |
| Open weights | Yes | Yes |
| Released | Apr 2026 | Dec 2025 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.