Provider guide
Where to run DeepSeek V4.1 Flash
33 live listings tracked — output prices vary 3.8× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.
FASTEST MEASURED
234 tok/s
measured throughput · $0.30 input / $1.20 output per million tokens
Sail Researchzero-retention$0.080 in/1M$0.40 out/1M49 tok/s1Mfp43 hours agoDekaLLMzero-retentionCHEAPEST$0.12 in/1M$0.40 out/1M43 tok/s1M—3 hours agoIo Netzero-retention$0.12 in/1M$0.41 out/1M73 tok/s1Mfp83 hours agoDeepInfrazero-retention through OpenRouterDirect and through OpenRouter$0.20 in/1M$0.14 in/1M$0.60 out/1Mdirect$0.42 out/1Mthrough OpenRouter75 tok/sthrough OpenRouter1Mfp83 hours agoMorphzero-retention$0.053 in/1M$0.43 out/1M60 tok/s1Mfp83 hours agoOpenInferencezero-retention$0.030 in/1M$0.50 out/1M7 tok/s1Mfp43 hours agoAtlasCloud$0.14 in/1M$0.56 out/1M111 tok/s1Mfp83 hours agoStreamLake$0.14 in/1M$0.56 out/1M75 tok/s1Mfp83 hours agoRekazero-retention$0.14 in/1M$0.56 out/1M51 tok/s1M—3 hours agoWaferzero-retention$0.050 in/1M$0.60 out/1M117 tok/s1M—3 hours agoInferenceNetzero-retention$0.040 in/1M$0.60 out/1M86 tok/s1M—3 hours agoDeepSeek$0.15 in/1M$0.60 out/1M82 tok/s1M—9 hours agoRelacezero-retention$0.026 in/1M$0.60 out/1M33 tok/s1M—3 hours agoCoreWeavezero-retention$0.20 in/1M$0.65 out/1M145 tok/s1Mfp89 hours agoFireworks AIzero-retention$0.22 in/1M$0.66 out/1M68 tok/s1M—3 hours agoDigitalOcean Gradientzero-retention$0.18 in/1M$0.72 out/1M53 tok/s1M—3 hours agoGMICloud$0.18 in/1M$0.72 out/1M98 tok/s1Mfp83 hours agoNextBitzero-retention$0.21 in/1M$0.84 out/1M78 tok/s1Mfp83 hours agoPhalazero-retention$0.21 in/1M$0.84 out/1M87 tok/s1M—3 hours agoIonstreamzero-retention$0.28 in/1M$1.15 out/1M157 tok/s1M—39 hours agoBaidu$0.30 in/1M$1.20 out/1M162 tok/s1Mfp83 hours agoParasailzero-retention$0.30 in/1M$1.20 out/1M131 tok/s1Mfp83 hours agoOpenRouter$0.30 in/1M$1.20 out/1M—1M—39 hours agoBasetenzero-retention$0.30 in/1M$1.20 out/1M79 tok/s1Mfp83 hours agoMakorazero-retention$0.30 in/1M$1.20 out/1M125 tok/s1Mfp83 hours agoNovita AIzero-retention through OpenRouterDirect and through OpenRouter$0.30 in/1M$1.20 out/1M78 tok/sthrough OpenRouter1Mfp83 hours agoAlibaba Cloud$0.30 in/1M$1.20 out/1M65 tok/s1M—15 hours agoSiliconFlowzero-retention$0.30 in/1M$1.20 out/1M93 tok/s1Mfp83 hours agoTogether AIzero-retention$0.30 in/1M$1.20 out/1M234 tok/s1M—3 hours agoModalzero-retention$0.30 in/1M$1.20 out/1M182 tok/s1M—3 hours agoVenice AIzero-retention$0.38 in/1M$1.50 out/1M101 tok/s1Mfp83 hours agoFireworks AIzero-retention$0.45 in/1M$1.80 out/1M173 tok/s1M—3 hours agoBasetenfast tier$0.60 in/1M$2.40 out/1M97 tok/s1Mfp323 hours ago
DeepInfra
Morph
Fireworks AI
DigitalOcean Gradient
Baidu
OpenRouter
Novita AIFull specs, hardware verdicts and benchmarks on the DeepSeek V4.1 Flash model page →