Head to head

DeepSeek V4 Flash (0731) vs Qwen3.5 9B

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).

DeepSeek V4 Flash (0731)Qwen3.5 9B
Provider Together AI Together AI
Input / MTok $0.14 $0.17
Output / MTok $0.28 $0.25
Cache read / MTok
Context window 1,000,000 262,144
Blended 1M in + 3M out $0.98 $0.92
Status ga ga

DeepSeek V4 Flash (0731) detail   Qwen3.5 9B detail