Every host for DeepSeek V3.1
| Host | Blended | Input | Output | Cached | Precision | Uptime |
|---|---|---|---|---|---|---|
| DeepInfra | $0.425 | $0.25 | $0.95 | $0.13 | fp4 | 100.0% |
| SiliconFlow | $0.453 | $0.27 | $1.00 | — | fp8 | 95.7% |
| CoreWeave | $0.825 | $0.55 | $1.65 | $0.55 | fp8 | 100.0% |
| SambaNova | $0.863 | $0.65 | $1.50 | — | fp8 | 97.4% |
| Mara | $0.875 | $0.60 | $1.70 | — | — | 0.0% |
Cheapest price over time
- Input
- Output
What your workload costs
- DeepInfra$170.00
- SiliconFlow$181.00
- CoreWeave$330.00
- SambaNova$345.00
- Mara$350.00
Rent GPUs or pay per token?
- Renting the GPUs
- $16,878/mo
- Break-even volume
- 39.7B tokens/mo
- Needed around the clock
- 15.1K tok/s
8 × H100 cost $16,878 a month. The cheapest API (DeepInfra) charges $0.425 per 1M blended tokens, so the GPUs only pay off above 39.7B tokens a month: the box must serve 15.1K 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.