TCO calculator
Total Cost of Ownership — the true monthly cost including GPU rental, idle time, and operational overhead. Compares apples-to-apples against API pricing.
Glossary →
Self-host vs API
For each open-weight model below: rental GPU options, single-stream throughput from public benchmarks, estimated monthly cost at your workload — and the cheapest API alternative for the same weights.
72B params
Qwen 2.5 72B Instruct
Qwen License
70B params
Llama 3.3 70B Instruct
Llama 3.3 Community License
47B params
Mixtral 8×7B Instruct
Apache 2.0
27B params
Gemma 2 27B Instruct
Gemma Terms of Use
14B params
Phi-4 14B
MIT
8B params
Llama 3.1 8B Instruct
Llama 3.1 Community License
7B params
Mistral 7B Instruct v0.3
Apache 2.0