Head to head

Qwen3.5 9B vs GPT-OSS 20B

On a blended workload of one million input tokens and three million output tokens, GPT-OSS 20B is 1.4× cheaper than Qwen3.5 9B.

Qwen3.5 9BGPT-OSS 20B
Provider Together AI Together AI
Input / MTok $0.17 $0.05
Output / MTok $0.25 $0.2
Cache read / MTok
Context window 262,144 128,000
Blended 1M in + 3M out $0.92 $0.65
Status ga ga

Qwen3.5 9B detail   GPT-OSS 20B detail