Official site · 🇨🇳 China
| Model | Intelligence | tok/s | $/M in | $/M out | $ / session | Cache |
|---|---|---|---|---|---|---|
| DeepSeek V4 Pro | 36.0 | 108.4 | $0.66 | $1.98 | $0.3103 | 96.9% |
| DeepSeek V4 Flash | 34.3 | — | $0.01 | $1.28 | $0.0475 | 94.9% |
| DeepSeek V3.2 | 21.5 | — | $0.28 | $0.42 | — | — |
| DeepSeek V3.1 Terminus | 14.8 | — | $0.27 | $1.00 | — | — |
| DeepSeek V3 | 8.5 | — | $0.26 | $1.03 | — | — |
| DeepSeek V3 0324 | 9.7 | — | $0.29 | $1.14 | — | — |
| DeepSeek R1 | 11.4 | — | $0.70 | $2.50 | — | — |
| DeepSeek V4 Flash Vision | 34.8 | 216.3 | $0.22 | $0.65 | — | — |
| DeepSeek V4.1 Flash | 39.5 | 209.5 | $0.30 | $1.20 | — | — |
2 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| DeepSeek V4.1 Flash | unknown | $0.15 | $0.60 | 1,048,576 | 100.0% |
| DeepSeek V4 Pro | unknown | $0.66 | $1.98 | 1,048,576 | 100.0% |
Everyone serving these models, cheapest first, with the numeric precision each one runs. The lab is often not the cheapest place to buy its own model.
| Model | Served by | Quant | $/M in | $/M out | Uptime 30m |
|---|---|---|---|---|---|
| DeepSeek V4 Pro | Baidu | fp8 | $0.13 | $0.40 | 100.0% |
| DeepSeek V4 Pro | Ionstream | unknown | $0.21 | $1.48 | 99.9% |
| DeepSeek V4 Pro | DeepSeek | unknown | $0.66 | $1.98 | 100.0% |
| DeepSeek V4 Pro | StreamLake | unknown | $0.66 | $1.98 | 98.5% |
| DeepSeek V4 Pro | Wafer | unknown | $0.33 | $4.20 | 100.0% |
| DeepSeek V4 Pro | Relace | fp4 | $0.90 | $2.70 | 100.0% |
| DeepSeek V4 Pro | Phala | unknown | $0.96 | $2.88 | — |
| DeepSeek V4 Pro | Novita | fp8 | $0.99 | $2.97 | 100.0% |
| DeepSeek V4 Pro | GMICloud | fp8 | $1.06 | $3.17 | 100.0% |
| DeepSeek V4 Pro | NextBit | fp8 | $1.06 | $3.17 | — |
| DeepSeek V4 Pro | DeepInfra | fp8 | $1.30 | $2.60 | 100.0% |
| DeepSeek V4 Pro | Alibaba / Qwen | unknown | $1.12 | $3.37 | 95.4% |
| DeepSeek V4 Pro | CoreWeave | fp8 | $1.31 | $3.96 | 100.0% |
| DeepSeek V4 Pro | AtlasCloud | fp8 | $1.32 | $3.96 | — |
| DeepSeek V4 Pro | Cloudflare | unknown | $1.32 | $3.96 | 99.9% |
| DeepSeek V4 Pro | DigitalOcean | unknown | $1.32 | $3.96 | — |
| DeepSeek V4 Pro | Parasail | fp8 | $1.32 | $3.96 | — |
| DeepSeek V4 Pro | SiliconFlow | fp8 | $1.32 | $3.96 | — |
| DeepSeek V4 Pro | Together | unknown | $1.32 | $3.96 | 97.9% |
| DeepSeek V4 Pro | Venice | unknown | $1.65 | $4.95 | — |
| DeepSeek V4 Flash | StreamLake | fp8 | $0.04 | $0.13 | 99.9% |
| DeepSeek V4 Flash | Sail Research | fp4 | $0.02 | $0.30 | 99.9% |
| DeepSeek V4 Flash | DeepInfra | fp8 | $0.06 | $0.18 | 100.0% |
| DeepSeek V4 Flash | Sail Research | fp4 | $0.02 | $0.42 | 99.9% |
| DeepSeek V4 Flash | Reka | unknown | $0.02 | $0.53 | 99.8% |
| DeepSeek V4 Flash | DigitalOcean | unknown | $0.12 | $0.24 | 100.0% |
| DeepSeek V4 Flash | BaseTen | fp8 | $0.13 | $0.26 | 99.9% |
| DeepSeek V4 Flash | BaseTen | fp8 | $0.13 | $0.26 | 100.0% |
| DeepSeek V4 Flash | CoreWeave | fp8 | $0.13 | $0.28 | 100.0% |
| DeepSeek V4 Flash | Cohere | unknown | $0.14 | $0.28 | 99.4% |
| DeepSeek V4 Flash | Parasail | fp8 | $0.14 | $0.28 | 99.8% |
| DeepSeek V4 Flash | Together | unknown | $0.14 | $0.28 | 99.6% |
| DeepSeek V4 Flash | Inceptron | fp4 | $0.05 | $0.65 | 100.0% |
| DeepSeek V4 Flash | Morph | bf16 | $0.14 | $0.40 | 100.0% |
| DeepSeek V4 Flash | Venice | unknown | $0.17 | $0.35 | 99.4% |
| DeepSeek V4 Flash | OpenInference | fp8 | $0.00 | $1.04 | 99.3% |
| DeepSeek V4 Flash | Wafer | unknown | $0.12 | $0.70 | 100.0% |
| DeepSeek V4 Flash | Mancer 2 | fp8 | $0.20 | $0.60 | 98.8% |
| DeepSeek V4 Flash | Relace | fp4 | $0.01 | $1.28 | 98.9% |
| DeepSeek V4 Flash | SiliconFlow | fp8 | $0.22 | $0.66 | 99.4% |
| DeepSeek V4 Flash | GMICloud | fp8 | $0.29 | $0.86 | 100.0% |
| DeepSeek V4 Flash | Phala | unknown | $0.31 | $0.92 | 100.0% |
| DeepSeek V4 Flash | Alibaba / Qwen | unknown | $0.35 | $1.06 | 99.1% |
| DeepSeek V4 Flash | NextBit | fp8 | $0.35 | $1.06 | 100.0% |
| DeepSeek V4 Flash | Novita | fp8 | $0.41 | $1.23 | 100.0% |
| DeepSeek V4 Flash | AtlasCloud | fp4 | $0.44 | $1.32 | 100.0% |
| DeepSeek V4 Flash | Baidu | fp8 | $0.44 | $1.32 | 100.0% |
| DeepSeek V4 Flash | Cloudflare | unknown | $0.44 | $1.32 | 100.0% |
| DeepSeek V3.2 | GMICloud | fp8 | $0.21 | $0.31 | 91.5% |
| DeepSeek V3.2 | AtlasCloud | fp8 | $0.26 | $0.38 | 96.5% |
| DeepSeek V3.2 | DeepInfra | fp4 | $0.26 | $0.38 | 100.0% |
| DeepSeek V3.2 | Venice | unknown | $0.27 | $0.39 | 100.0% |
| DeepSeek V3.2 | SiliconFlow | fp8 | $0.26 | $0.42 | 99.6% |
| DeepSeek V3.2 | Baidu | fp8 | $0.28 | $0.42 | 100.0% |
| DeepSeek V3.2 | DigitalOcean | unknown | $0.30 | $0.96 | 98.1% |
| DeepSeek V3.2 | Alibaba / Qwen | fp8 | $0.37 | $1.11 | 96.9% |
| DeepSeek V3.2 | Friendli | unknown | $0.50 | $1.50 | 100.0% |
| DeepSeek V3.2 | unknown | $0.56 | $1.68 | 99.7% | |
| DeepSeek V3.2 | Phala | unknown | $1.00 | $1.00 | 98.3% |
| DeepSeek V3.2 | Mara | unknown | $3.00 | $4.50 | — |
| DeepSeek V3.2 | SambaNova | unknown | $3.00 | $4.50 | 98.7% |
| DeepSeek V3.1 Terminus | SiliconFlow | fp8 | $0.27 | $1.00 | 99.4% |
| DeepSeek V3.1 Terminus | AtlasCloud | fp8 | $0.30 | $1.00 | 99.4% |
| DeepSeek V3.1 Terminus | StreamLake | unknown | $0.34 | $1.03 | 99.1% |
| DeepSeek V3 | StreamLake | unknown | $0.26 | $1.03 | 99.2% |
| DeepSeek V3 | DeepInfra | fp4 | $0.32 | $0.89 | 92.7% |
| DeepSeek V3 0324 | SiliconFlow | fp8 | $0.25 | $1.00 | 99.9% |
| DeepSeek V3 0324 | GMICloud | fp8 | $0.29 | $1.14 | 99.1% |
| DeepSeek R1 | Novita | fp8 | $0.70 | $2.50 | 100.0% |
| DeepSeek V4 Flash Vision | DeepInfra | fp8 | $0.22 | $0.65 | 99.9% |
| DeepSeek V4 Flash Vision | GMICloud | fp8 | $0.44 | $1.32 | 100.0% |
| DeepSeek V4 Flash Vision | Novita | unknown | $0.44 | $1.32 | 100.0% |
| DeepSeek V4 Flash Vision | SiliconFlow | fp8 | $0.44 | $1.32 | 100.0% |
| DeepSeek V4.1 Flash | Morph | fp8 | $0.02 | $0.38 | 99.7% |
| DeepSeek V4.1 Flash | Decart | fp4 | $0.09 | $0.18 | 100.0% |
| DeepSeek V4.1 Flash | Sail Research | fp4 | $0.08 | $0.40 | 99.9% |
| DeepSeek V4.1 Flash | Relace | unknown | $0.02 | $0.60 | 99.9% |
| DeepSeek V4.1 Flash | InferenceNet | unknown | $0.04 | $0.60 | 99.8% |
| DeepSeek V4.1 Flash | OpenInference | fp4 | $0.01 | $0.68 | 94.4% |
| DeepSeek V4.1 Flash | Wafer | unknown | $0.05 | $0.60 | 100.0% |
| DeepSeek V4.1 Flash | DekaLLM | unknown | $0.12 | $0.40 | 99.7% |
| DeepSeek V4.1 Flash | DeepInfra | fp8 | $0.14 | $0.42 | 99.9% |
| DeepSeek V4.1 Flash | Baidu | fp8 | $0.14 | $0.56 | 99.0% |
| DeepSeek V4.1 Flash | AtlasCloud | fp8 | $0.14 | $0.56 | 99.0% |
| DeepSeek V4.1 Flash | DeepSeek | unknown | $0.15 | $0.60 | 100.0% |
| DeepSeek V4.1 Flash | CoreWeave | fp8 | $0.20 | $0.65 | 99.2% |
| DeepSeek V4.1 Flash | StreamLake | fp8 | $0.18 | $0.72 | 98.9% |
| DeepSeek V4.1 Flash | DigitalOcean | unknown | $0.18 | $0.72 | 100.0% |
| DeepSeek V4.1 Flash | GMICloud | fp8 | $0.18 | $0.72 | 99.8% |
| DeepSeek V4.1 Flash | Ionstream | unknown | $0.10 | $1.15 | 100.0% |
| DeepSeek V4.1 Flash | NextBit | fp8 | $0.21 | $0.84 | 100.0% |
| DeepSeek V4.1 Flash | Phala | unknown | $0.21 | $0.84 | 99.3% |
| DeepSeek V4.1 Flash | Makora | fp8 | $0.20 | $0.99 | 99.9% |
| DeepSeek V4.1 Flash | Novita | fp8 | $0.24 | $0.96 | 95.6% |
| DeepSeek V4.1 Flash | Alibaba / Qwen | unknown | $0.30 | $1.20 | 100.0% |
| DeepSeek V4.1 Flash | BaseTen | fp8 | $0.30 | $1.20 | 99.8% |
| DeepSeek V4.1 Flash | BaseTen | fp8 | $0.30 | $1.20 | 99.9% |
| DeepSeek V4.1 Flash | Fireworks | unknown | $0.30 | $1.20 | 99.9% |
| DeepSeek V4.1 Flash | Modal | unknown | $0.30 | $1.20 | 100.0% |
| DeepSeek V4.1 Flash | Parasail | fp8 | $0.30 | $1.20 | 99.9% |
| DeepSeek V4.1 Flash | SiliconFlow | fp8 | $0.30 | $1.20 | 99.9% |
| DeepSeek V4.1 Flash | Together | unknown | $0.30 | $1.20 | 100.0% |
| DeepSeek V4.1 Flash | Venice | fp8 | $0.38 | $1.50 | 99.9% |
| DeepSeek V4.1 Flash | Fireworks | unknown | $0.45 | $1.80 | 100.0% |