Every host for Llama-4-Maverick-17B-128E-Instruct-FP8
| Host | Blended | Input | Output | Cached | Precision | Uptime |
|---|---|---|---|---|---|---|
| Lambda | $0.063 | $0.05 | $0.10 | — | — | — |
| DeepInfra | $0.35 | $0.20 | $0.80 | — | — | — |
| Novita | $0.415 | $0.27 | $0.85 | — | — | — |
| Azure | $0.438 | $0.25 | $1.00 | — | — | — |
| Watsonx | $0.649 | $0.371 | $1.48 | — | — | — |
Cheapest price over time
- Input
- Output
What your workload costs
- Lambda$25.00
- DeepInfra$140.00
- Novita$166.00
- Azure$175.00
- Watsonx$259.70
Rent GPUs or pay per token?
- Renting the GPUs
- $16,878/mo
- Break-even volume
- 270B tokens/mo
- Needed around the clock
- 103K tok/s
8 × H100 cost $16,878 a month. The cheapest API (Lambda) charges $0.063 per 1M blended tokens, so the GPUs only pay off above 270B tokens a month: the box must serve 103K 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 Meta
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.