Every host for DeepSeek V4 Pro 0813
| Host | Blended | Input | Output | Cached | Precision | Uptime |
|---|---|---|---|---|---|---|
| Sail Research | $0.538 | $0.45 | $0.80 | $0.065 | fp4 | 100.0% |
| Streamlake | $0.99 | $0.66 | $1.98 | $0.022 | — | 100.0% |
| Ionstream | $1.01 | $0.37 | $2.93 | $0.088 | — | 100.0% |
| Sail | $1.38 | $0.92 | $2.77 | $0.04 | — | — |
| Phala | $1.44 | $0.957 | $2.88 | $0.099 | — | 100.0% |
| Wafer | $1.44 | $0.25 | $5.00 | $0.249 | mxfp4 | 99.4% |
| Novita | $1.48 | $0.99 | $2.97 | $0.033 | fp8 | 100.0% |
| GMI Cloud | $1.58 | $1.06 | $3.17 | $0.035 | fp8 | 100.0% |
| Dekallm | $1.61 | $0.88 | $3.80 | $0.15 | — | 99.3% |
| DeepInfra | $1.63 | $1.30 | $2.60 | $0.10 | fp8 | 99.9% |
| Alibaba Cloud | $1.68 | $1.12 | $3.37 | $0.112 | — | 100.0% |
| CoreWeave | $1.97 | $1.31 | $3.96 | $0.044 | fp8 | 99.0% |
| Weights & Biases | $1.97 | $1.31 | $3.96 | $0.044 | — | — |
| Atlas Cloud | $1.98 | $1.32 | $3.96 | $0.132 | fp8 | 100.0% |
| Baidu | $1.98 | $1.32 | $3.96 | $0.042 | fp8 | 100.0% |
| Baseten | $1.98 | $1.32 | $3.96 | $0.132 | — | — |
| Cloudflare | $1.98 | $1.32 | $3.96 | $0.044 | — | 99.5% |
| DeepSeek | $1.98 | $1.32 | $3.96 | $0.044 | — | 100.0% |
| Digitalocean | $1.98 | $1.32 | $3.96 | $0.044 | — | 94.6% |
| Fireworks | $1.98 | $1.32 | $3.96 | $0.044 | — | — |
| Nebius | $1.98 | $1.32 | $3.96 | — | — | — |
| Parasail | $1.98 | $1.32 | $3.96 | $0.044 | fp8 | 98.6% |
| Perplexity | $1.98 | $1.32 | $3.96 | $0.044 | — | — |
| SiliconFlow | $1.98 | $1.32 | $3.96 | $0.044 | fp8 | 100.0% |
| Tencent Cloud | $1.98 | $1.32 | $3.96 | $0.044 | — | — |
| Together | $1.98 | $1.32 | $3.96 | $0.13 | — | 98.6% |
| Relace | $2.17 | $0.225 | $8.00 | $0.23 | fp4 | 99.8% |
| Venice | $2.47 | $1.65 | $4.95 | $0.165 | — | 98.4% |
| Inferenceapis | $2.58 | $1.72 | $5.15 | $0.17 | — | — |
Cheapest price over time
- Input
- Output
What your workload costs
- Sail Research$215.00
- Streamlake$396.00
- Ionstream$404.00
- Sail$553.00
- Phala$574.86
- Wafer$575.00
- Novita$594.00
- GMI Cloud$633.60
- Dekallm$644.00
- DeepInfra$650.00
- Alibaba Cloud$673.20
- CoreWeave$789.00
17 more hosts cost more.
Rent GPUs or pay per token?
- Renting the GPUs
- $16,878/mo
- Break-even volume
- 31.4B tokens/mo
- Needed around the clock
- 11.9K tok/s
8 × H100 cost $16,878 a month. The cheapest API (Sail Research) charges $0.538 per 1M blended tokens, so the GPUs only pay off above 31.4B tokens a month: the box must serve 11.9K tokens a second, every second.
GPU price: cheapest on-demand offer per GPU-hour × 730 hours. No throughput is assumed: it depends on the model size, precision, batch size and serving software, so measure yours. Power, storage, egress and engineering time are not included.
More from DeepSeek
Quality data: LMArena (CC BY 4.0), Epoch AI (CC BY 4.0). Prices: OpenRouter, LiteLLM (MIT), provider pages. List prices in US dollars before discounts and taxes; other currencies converted at the European Central Bank reference rate. The figures on these pages are free to reuse under CC BY 4.0 with a link to this page.