> ## Documentation Index > Fetch the complete documentation index at: https://docs.gcore.com/llms.txt > Use this file to discover all available pages before exploring further. # Deploy a model from the Application Catalog The [Application Catalog](/edge-ai/everywhere-inference/application-catalog) provides pre-built open-source AI models that you can deploy without building or configuring a container image. [Request a quota increase](/edge-ai/everywhere-inference/quotas/request-quota-increase) if the account quota is insufficient for the selected flavor. ## Create a deployment Complete the following steps to select an application, configure compute resources, and deploy the model. ### Step 1. Open the Create Deployment form There are two ways to open the **Create Deployment** form: * Navigate to **Everywhere Inference** > **Deployments** and click **Deploy application from catalog**. * Open a model detail page in the Application Catalog and click **Deploy Application**. The form opens with the selected model pre-filled. Deployments page showing the Deploy application from catalog and Deploy custom inference buttons ### Step 2. Select an application Under **Deployment Configuration**, click the **Application** dropdown and select the model to deploy. Create Deployment form — Deployment Configuration and Routing placement sections ### Step 3. Select regions Under **Routing placement**, click **Select region** and select up to six regions where the model will run. ### Step 4. Configure application modules Under **Application modules**, configure the main module: * **Expose** — controls whether the module endpoint is publicly accessible. Enabled by default. * **Flavor type** — select **CPU-optimized** or **GPU-optimized** depending on the model requirements. * **Flavor** — select the hardware configuration from the dropdown. * **Minimum pods** — the minimum number of pods to keep running during low-traffic periods. * **Maximum pods** — the maximum number of pods Gcore provisions during peak traffic. Under **Parameters**, adjust inference settings for the model: * **Chunked prefill** — splits long prompts into smaller chunks for processing. Default: `true`. * **CPU Cache** — caches KV tensors in CPU memory when GPU memory is full. Default: `true`. * **Prefix caching** — reuses cached KV state for repeated prompt prefixes. Default: `true`. Application modules section showing Parameters and Open WebUI module The main application module cannot be removed. Some models include optional modules listed below the main module card. **Open WebUI** is a browser-based chat interface that lets users interact with the deployed model without writing code. To enable it, toggle it on. When enabled, Open WebUI runs as a separate module with its own compute settings: * **Expose** — controls whether the Open WebUI endpoint is publicly accessible. * **Flavor type** — select **CPU-optimized** or **GPU-optimized**. * **Flavor** — select the hardware configuration for the Open WebUI module. * **Minimum pods** and **Maximum pods** — set the pod scaling range for the module. Open WebUI module enabled showing Expose, Flavor type, Flavor, and pod count settings Each enabled module consumes one inference instance from the quota. ### Step 5. Name the deployment Under **Deployment details**, enter a name for the deployment. Use only letters and numbers — hyphens are not allowed in deployment names. ### Step 6. Set additional options (Optional) Enable the **Enable API Key authentication** toggle to restrict access using [API keys](/edge-ai/everywhere-inference/api-keys/create-inference-deployment-with-auth). Additional options section with Enable API Key authentication toggle enabled and API key selector ### Step 7. Deploy Review the estimated cost in the right panel, then click **Deploy model**. Gcore creates the deployment and opens the **Deployments** page, where the deployment status is visible. ## Deployment status The new deployment appears on the **Deployments** page with a **Deploying** status. Once all pods are running, the status changes to **Active**. The endpoint URL becomes available on the deployment detail page. Use it to send inference requests as described in [Query a deployed model](/edge-ai/everywhere-inference/ai-models/query-deployed-model). Logs, compute settings, and other deployment options are available on the [deployment detail page](/edge-ai/everywhere-inference/ai-models/manage-ai-model-deployments).