Every host for Hy4 preview
| Host | Blended | Input | Output | Cached | Precision | Uptime |
|---|---|---|---|---|---|---|
| Cohere | $1.25 | $0.834 | $2.50 | $0.042 | — | 97.3% |
| DeepInfra | $1.25 | $0.834 | $2.50 | $0.042 | fp4 | 99.8% |
| Novita | $1.25 | $0.834 | $2.50 | $0.042 | fp8 | 99.8% |
| SiliconFlow | $1.25 | $0.834 | $2.50 | $0.042 | fp8 | 100.0% |
| Tencent Cloud | $1.25 | $0.834 | $2.50 | $0.042 | fp8 | 99.4% |
Cheapest price over time
- Input
- Output
What your workload costs
- Cohere$500.30
- DeepInfra$500.30
- Novita$500.30
- SiliconFlow$500.30
- Tencent Cloud$500.30
Rent GPUs or pay per token?
- Renting the GPUs
- $16,878/mo
- Break-even volume
- 13.5B tokens/mo
- Needed around the clock
- 5.1K tok/s
8 × H100 cost $16,878 a month. The cheapest API (Cohere) charges $1.25 per 1M blended tokens, so the GPUs only pay off above 13.5B tokens a month: the box must serve 5.1K 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 Tencent Cloud
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.