Head to head

Ministral 3 8B vs Qwen3.5 9B

On a blended workload of one million input tokens and three million output tokens, Ministral 3 8B is 1.5× cheaper than Qwen3.5 9B.

Ministral 3 8BQwen3.5 9B
Provider Mistral Together AI
Input / MTok $0.15 $0.17
Output / MTok $0.15 $0.25
Cache read / MTok
Context window 262,144
Blended 1M in + 3M out $0.60 $0.92
Status ga ga

Ministral 3 8B detail   Qwen3.5 9B detail