Provider guide
Where to run Qwen3.6 35B A3B
11 live listings tracked — output prices vary 2.6× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
140 tok/s
measured throughput · $0.25 input / $1.25 output per million tokens
Darkbloom$0.050 in/1M$0.70 out/1M74 tok/s262Kfp42 hours agoAkashMLzero-retention$0.10 in/1M$0.90 out/1M44 tok/s262Kfp82 hours agoDeepInfrazero-retention through OpenRouterDirect and through OpenRouter$0.10 in/1M$0.95 out/1M30 tok/sthrough OpenRouter262Kfp88 hours ago direct14 hours ago through OpenRouterOpenRouterCHEAPEST$0.15 in/1M$1.00 out/1M—262K—2 hours agoVenice AIzero-retention$0.10 in/1M$1.00 out/1M98 tok/s256Kfp82 hours agoParasailzero-retention$0.15 in/1M$1.00 out/1M30 tok/s262Kfp82 hours agoAtlasCloud$0.19 in/1M$1.11 out/1M95 tok/s262Kfp82 hours agoCoreWeavezero-retention$0.25 in/1M$1.25 out/1M140 tok/s262Kfp82 hours agoPhalazero-retention$0.20 in/1M$1.27 out/1M89 tok/s262K—2 hours agoNovita AI$0.25 in/1M$1.49 out/1M—262K—8 hours agoSiliconFlowzero-retention$0.24 in/1M$1.80 out/1M90 tok/s262Kfp82 hours ago
DeepInfra
Novita AIFull specs, hardware verdicts and benchmarks on the Qwen3.6 35B A3B model page →