Head to head
Qwen3.5 9B vs DeepSeek V4 Flash (0731)
On a blended workload of one million input tokens and three million output tokens, Qwen3.5 9B is 1.1× cheaper than DeepSeek V4 Flash (0731).
| Qwen3.5 9B | DeepSeek V4 Flash (0731) | |
|---|---|---|
| Provider | Together AI | Together AI |
| Input / MTok | $0.17 | $0.14 |
| Output / MTok | $0.25 | $0.28 |
| Cache read / MTok | — | — |
| Context window | 262,144 | 1,000,000 |
| Blended 1M in + 3M out | $0.92 | $0.98 |
| Status | ga | ga |