> For the complete documentation index, see [llms.txt](https://doc.faceplugin.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.faceplugin.com/integration/deploy-and-host/production-deployment.md).

# Production deployment

Production checklist for Faceplugin server SDKs. TLS, health checks, activate, no Gradio, networking, scale, and upgrades.

Use this checklist after a first-run install works. These are **how you run** the product. They are not new product features.

## Before you go live

1. Use a Docker Hub image, or a **fixed image tag** you control. Avoid surprise `latest` changes in production if you need a locked build.
2. Do **not** run Gradio (`demo` / `demo.py`) in production. Gradio is a local test UI only.
3. Activate with `POST /api/activate`. Do not depend on an interactive terminal. Detached Compose has no TTY.
4. Point your load balancer or orchestrator at `GET /api/health`. No license is required for health.
5. Put the APIs behind **your** reverse proxy. End TLS at nginx, Caddy, or your cloud load balancer. Open only the ports you need.
6. Keep the APIs on your private network when you can. The Flask apps allow CORS from any origin (`*`). Do not treat CORS as security.
7. Run **one container or process per product**. Prefer **combined** Face Recognition + Liveness (**8083**) and Document Reader with authenticity (**8082**) instead of separate liveness-only services. To scale, add more containers. Follow the license and machine-code rules on each platform page.
8. Read logs from `docker logs` or process stdout. Faceplugin does not collect cloud telemetry from these apps.

## License and machine code

A **machine code** identifies the server or container for licensing.

{% hint style="warning" %}
A Docker container and the Linux host have **different** machine codes. If you run in Docker, request the license with the code from the **container**.
{% endhint %}

On Linux, mounting `/etc/machine-id` into each container can let several containers share one machine code. On Docker Desktop, omit that volume. Each container may need its own license. Details: [Request a License & Support](/help/request-a-license-and-support.md).

## Upgrade path

1. Pull the new image, or replace files in `lib/cpu/`.
2. Restart the container or process.
3. Activate again if the machine code changed.
4. Smoke-test `GET /api/health` and one process endpoint (for example `/api/detect` or `/api/documentRecognition`).

## What production is not

These surfaces are for evaluation only:

* [Playground](https://playground.faceplugin.com/)
* [Hugging Face Spaces](https://huggingface.co/Faceplugin-Ltd)
* Gradio demos on ports 9002–9006

Do not point production traffic at them.

## Quick links

| Topic                     | Page                                                                         |
| ------------------------- | ---------------------------------------------------------------------------- |
| Topology and what you own | [Architecture](/integration/deploy-and-host/architecture.md)                 |
| CPU, RAM, shm, ports      | [Hosting requirements](/integration/deploy-and-host/hosting-requirements.md) |
| Endpoint contracts        | [HTTP API](/integration/http-api.md)                                         |
| First-run curls           | [Try it](/getting-started/try-it.md)                                         |

### Related documentation

* [Deploy and host](/integration/deploy-and-host.md) · [Troubleshooting](/help/resources/troubleshooting.md) · [Status codes](/help/resources/status-codes.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://doc.faceplugin.com/integration/deploy-and-host/production-deployment.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.