Every host for Nemotron 3.5 Super VL
| Host | Blended | Input | Output | Cached | Precision | Uptime |
|---|---|---|---|---|---|---|
| Darkbloom | $0.133 | $0.06 | $0.35 | $0.01 | nvfp4 | 100.0% |
| Crusoe | $0.175 | $0.10 | $0.40 | $0.08 | bf16 | 99.6% |
| CoreWeave | $0.238 | $0.15 | $0.50 | $0.02 | nvfp4 | 99.3% |
| DeepInfra | $0.238 | $0.15 | $0.50 | $0.02 | fp4 | 99.4% |
| Io Net | $0.238 | $0.15 | $0.50 | $0.02 | nvfp4 | 99.5% |
| Together | $0.238 | $0.15 | $0.50 | $0.02 | — | 100.0% |
| Baseten | $0.275 | $0.20 | $0.50 | $0.04 | fp4 | 100.0% |
| Venice | $0.50 | $0.25 | $1.25 | $0.125 | fp4 | 100.0% |
Cheapest price over time
- Input
- Output
What your workload costs
- Darkbloom$53.00
- Crusoe$70.00
- CoreWeave$95.00
- DeepInfra$95.00
- Io Net$95.00
- Together$95.00
- Baseten$110.00
- Venice$200.00
Rent GPUs or pay per token?
- Renting the GPUs
- $16,878/mo
- Break-even volume
- 127B tokens/mo
- Needed around the clock
- 48.5K tok/s
8 × H100 cost $16,878 a month. The cheapest API (Darkbloom) charges $0.133 per 1M blended tokens, so the GPUs only pay off above 127B tokens a month: the box must serve 48.5K 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 NVIDIA
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.