Provider guide
Where to run Qwen3 Next 80B A3B Instruct
6 live listings tracked — output prices vary 1.9× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
56 tok/s
measured throughput · $0.098 input / $0.78 output per million tokens
Alibaba Cloud$0.098 in/1M$0.78 out/1M56 tok/s131K—3 hours agoDeepInfrazero-retention through OpenRouterDirect and through OpenRouter$0.090 in/1M$1.10 out/1M46 tok/sthrough OpenRouter262Kfp89 hours ago direct3 hours ago through OpenRouterOpenRouterCHEAPEST$0.10 in/1M$1.10 out/1M—262K—3 hours agoParasailzero-retention$0.10 in/1M$1.10 out/1M17 tok/s262Kfp83 hours agoGoogle Vertex AIzero-retention$0.15 in/1M$1.20 out/1M13 tok/s262K—3 hours agoNovita AIzero-retention through OpenRouterDirect and through OpenRouter$0.15 in/1M$1.50 out/1M8 tok/sthrough OpenRouter131Kbf169 hours ago direct3 hours ago through OpenRouter
DeepInfra
Google Vertex AI
Novita AIFull specs, hardware verdicts and benchmarks on the Qwen3 Next 80B A3B Instruct model page →