> See the full index of documentation pages at https://docs.zettabyte.com/llms.txt.

# Deploy and serve

> Deploy models as dedicated inference endpoints, call them over the inference API, and track token usage.

Source: https://docs.zettabyte.com/docs/deploy-and-serve

Last updated: 2026-10-17

## Pages

- [GPU memory calculator for LLMs](https://docs.zettabyte.com/docs/deploy-and-serve/gpu-memory-calculator.md): Estimate LLM memory requirements: whether a model fits in GPU memory, and what to change when it runs out of memory.
- [Deploy a model](https://docs.zettabyte.com/docs/deploy-and-serve/deploy-a-model.md): Deploy a model as a dedicated inference endpoint.
- [Call your endpoint](https://docs.zettabyte.com/docs/deploy-and-serve/call-your-endpoint.md): Send requests to a deployed inference endpoint over HTTP.
- [Token usage](https://docs.zettabyte.com/docs/deploy-and-serve/token-usage.md): Track inference token consumption and cost.

## Optional

- [Full Deploy and serve documentation](https://docs.zettabyte.com/docs/deploy-and-serve/llms-full.txt): every Deploy and serve page as one markdown file