Provider guide
Where to run Qwen3 Coder 480B A35B
10 live offers tracked — output prices vary 4.9× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
69 tok/s
measured throughput · $0.30 input / $1.00 output per million tokens
OpenRouterCHEAPEST$0.30 in/1M$1.00 out/1M—262K—15m agoDeepInfra$0.30 in/1M$1.00 out/1M32 tok/s262Kfp46d agoNovita AI$0.38 in/1M$1.55 out/1M—262K—6h agoAlibaba Cloud$0.97 in/1M$4.88 out/1M18 tok/s262Kfp813m agoAlibaba Cloud$0.97 in/1M$4.88 out/1M45 tok/s262Kfp86d agoDeepInfraturbo tier$0.30 in/1M$1.00 out/1M69 tok/s262Kfp413m agoCoreWeavezero-retention$1.00 in/1M$1.50 out/1M45 tok/s262Kbf1613m agoVenice AIzero-retention$0.35 in/1M$1.50 out/1M37 tok/s256Kfp813m agoNovita AIzero-retention$0.38 in/1M$1.55 out/1M14 tok/s262Kfp813m agoGoogle Vertex AIzero-retention$0.22 in/1M$1.80 out/1M25 tok/s262K—13m ago
Novita AI
Google Vertex AIFull specs, hardware verdicts and benchmarks on theQwen3 Coder 480B A35B model page →