Tools
Calculators we use when we size and price a private LLM. They run in your browser; nothing you enter is sent anywhere.
- LLM VRAM calculator
GPU memory for a model's weights and KV cache, by model size, precision, context length and number of users, with the GPUs that fit.
- LLM cost calculator: API vs self-hosted
Monthly cost of your token volume on an API, a rented GPU server and owned hardware, and the volume where self-hosting breaks even.