»Model API
Per-token pricing
| Model | Context | Input / 1M | Output / 1M |
|---|---|---|---|
| Llama 4 Maverick | 1M | $0.18 | $0.6 |
| DeepSeek V3.2 | 128K | $0.27 | $1.1 |
| Qwen3 235B | 256K | $0.2 | $0.8 |
| Mistral Large 3 | 128K | $2 | $6 |
| GPT-OSS 120B | 128K | $0.15 | $0.6 |
| Kimi K2 | 256K | $0.6 | $2.5 |
»GPU Cloud
Hourly GPU rates
| GPU | Memory | On-demand | Reserved |
|---|---|---|---|
| B200 | 192 GB HBM3e | $4.99/hr | $3.79/hr |
| H200 | 141 GB HBM3e | $3.49/hr | $2.69/hr |
| H100 | 80 GB HBM3 | $2.39/hr | $1.89/hr |
| A100 | 80 GB HBM2e | $1.29/hr | $0.99/hr |
| L40S | 48 GB GDDR6 | $0.89/hr | $0.69/hr |