Cost calculator
Estimate monthly inference cost across every tracked model. Tweak token counts and the table re-sorts. Cache pricing and batch discounts are applied when supported by the model.
| Model | Per request | Per month |
|---|---|---|
| Llama 3.3 70B | $0.0015 | $155 |
| DeepSeek V3 | $0.0024 | $245 |
| GPT-5 mini | $0.0032 | $325 |
| Gemini 2.5 Flash | $0.0040 | $400 |
| Qwen3 235B | $0.0040 | $400 |
| DeepSeek R1 | $0.0049 | $494 |
| Claude Haiku 4.5 | $0.010 | $1000 |
| Mistral Large 2 | $0.016 | $1600 |
| Gemini 2.5 Pro | $0.016 | $1625 |
| GPT-5 | $0.016 | $1625 |
| GPT-4.1 | $0.018 | $1800 |
| GPT-4o | $0.022 | $2250 |
| Claude Sonnet 4.6 | $0.030 | $3000 |
| Grok 4 | $0.030 | $3000 |
| Claude Opus 4.7 | $0.150 | $15.0K |