Provider guide
Where to run Qwen3 VL 235B A22B Instruct
6 live listings tracked — output prices vary 2.2× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
45 tok/s
measured throughput · $0.30 input / $1.50 output per million tokens
DeepInfrazero-retention through OpenRouterDirect and through OpenRouter$0.20 in/1M$0.88 out/1M8 tok/sthrough OpenRouter262Kfp82 hours agoAlibaba Cloud$0.26 in/1M$1.04 out/1M39 tok/s131K—2 hours agoNovita AIzero-retention through OpenRouterDirect and through OpenRouter$0.30 in/1M$1.50 out/1M45 tok/sthrough OpenRouter131Kbf162 hours agoParasailzero-retention$0.21 in/1M$1.90 out/1M22 tok/s131Kfp82 hours agoVenice AIzero-retention$0.21 in/1M$1.90 out/1M33 tok/s128Kfp82 hours agoOpenRouterCHEAPEST$0.21 in/1M$1.90 out/1M—262K—2 hours ago
DeepInfraFull specs, hardware verdicts and benchmarks on the Qwen3 VL 235B A22B Instruct model page →