Provider guide
Where to run Qwen3 Next 80B A3B Instruct
9 live offers 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
86 tok/s
measured throughput · $0.098 input / $0.78 output per million tokens
Alibaba Cloud$0.098 in/1M$0.78 out/1M86 tok/s131K—7d agoAlibaba Cloud$0.098 in/1M$0.78 out/1M83 tok/s131Kfp842m agoOpenRouterCHEAPEST$0.090 in/1M$1.10 out/1M—262K—43m agoDeepInfra$0.090 in/1M$1.10 out/1M—262Kfp87h agoNovita AI$0.15 in/1M$1.50 out/1M—131K—7h agoDeepInfrazero-retention$0.090 in/1M$1.10 out/1M66 tok/s262Kfp842m agoParasailzero-retention$0.10 in/1M$1.10 out/1M67 tok/s262Kfp842m agoGoogle Vertex AIzero-retention$0.15 in/1M$1.20 out/1M78 tok/s262K—42m agoNovita AIzero-retention$0.15 in/1M$1.50 out/1M69 tok/s131Kbf1642m ago
DeepInfra
Novita AI
Google Vertex AIFull specs, hardware verdicts and benchmarks on theQwen3 Next 80B A3B Instruct model page →