Deploying Open WebUI with Scaleway's Generative APIs

Integrating Scaleway's Generative APIs with Open WebUI allows you to leverage powerful AI models in a user-friendly, self-hosted interface. Open WebUI is an open-source frontend that simplifies interactions with AI models, providing an intuitive alternative to proprietary AI interfaces.

This tutorial will guide you through:

• Configuring Generative APIs on Scaleway
• Installing Open WebUI locally
• Deploying Open WebUI using Scaleway's Serverless Containers

By the end, you will have a fully functional AI-powered setup.

Before you start

To complete the actions presented below, you must have:

Prerequisites

• A Scaleway account logged into the console
• Owner status or appropriate IAM permissions
• Docker installed on your system

Configuring Generative APIs on Scaleway

Scaleway's Generative APIs provide access to advanced AI models via simple API calls. Follow the instructions below to set up Generative APIs

Creating an API key

1. Log in to the Scaleway console.
2. Click on your Organization name in the top right corner of the console to display the menu.
3. Click API keys in the pop-up menu.
4. Click Generate a New API Key, provide a name, and store the key securely.

Identifying the API endpoint

Scaleway assigns a unique endpoint for your Generative API access. This can be found in the Generative APIs section of the console. The format typically follows:

https://api.scaleway.ai/<YOUR_PROJECT_ID>/v1

If you have not created a dedicated Project and are using the default Project created with your Scaleway account, you can use the following URL: https://api.scaleway.ai/v1. You can also find the exact URL snippet including your project_id by visiting the Generative API Playground and clicking View Code.

Installing Open WebUI locally

Open WebUI provides an easy-to-use interface for interacting with AI models. The recommended installation method is via Docker.

Running Open WebUI with Scaleway integration

1. Pull and run the Docker container:

   docker run -p 3000:8080 \
     -e OPENAI_API_BASE_URL=https://api.scaleway.ai/<YOUR_PROJECT_ID>/v1 \
     -e OPENAI_API_KEY=<YOUR_API_KEY> \
     -v open-webui:/app/backend/data \
     --name open-webui \
     --restart always \
     ghcr.io/open-webui/open-webui:main

   Note

   Replace <YOUR_PROJECT_ID> and <YOUR_API_KEY> with your actual values.

2. Navigate to http://localhost:3000 in your browser to access the Open WebUI interface.

3. Click Sign Up to create an administrator account.

4. Select a model from the available Scaleway AI models and start interacting.

For more details, refer to the Open WebUI Quick Start Guide.

Deploying Open WebUI using Scaleway's Serverless Containers (Optional)

For cloud-based deployment, Scaleway's Serverless Containers offer a scalable solution.

Deploying on Scaleway via the console

1. In the Scaleway console, navigate to Serverless > Containers.
2. Click Deploy Container to display the container creation wizard.
3. Select an external registry and use the ghcr.io/open-webui/open-webui:main image.

4. Enter a name for your container and configure its resources.

5. Click Advanced options and configure secrets:

   • Set OPENAI_API_BASE_URL and OPENAI_API_KEY with your specific values.

6. Click Deploy container to launch deployment. Once deployed, note the provided endpoint to access Open WebUI.

   Note

   Since the Open WebUI default image is around 5 GB size, deployment can take a couple of minutes before being ready.

Deploying on Scaleway using the ClI

Ensure you have a Scaleway Serverless Containers namespace

Scaleway Serverless Containers are deployed inside a namespace. Run the following command to create a namespace:

scw container namespace create name=openwebui-ns region=fr-par

Run the following command to retrieve the namespace ID:

scw container namespace list

Create the Serverless Container

Run the following command to create the container with your Docker image and secret injection:

$ scw container container create name=open-webui \
    namespace-id=<YOUR_NAMESPACE_ID> \
    registry-image="ghcr.io/open-webui/open-webui:0.5.10" \
    environment-variables.OPENAI_API_BASE_URL="https://api.scaleway.ai/<YOUR_PROJECT_ID>/v1" \
    secret-environment-variables.0.key="OPENAI_API_KEY" \
    secret-environment-variables.0.value="<YOUR_SECRET_KEY>" \
    region=fr-par \
    cpu-limit=1000 \
    memory-limit=4000 \
    local-storage-limit=10000 \
    timeout=600s

For more information, refer to Scaleway's Serverless Containers documentation.