Deploy Your First Model in 5 Minutes

This guide walks you through deploying your first AI model on Mycelis — from account creation to a live API endpoint, ready in just a few minutes.

Prerequisites

Before you deploy, make sure three things are in place.

1. A Mycelis account

If you haven't signed up yet, go to mycelis.ai and create a free account. Registration takes less than a minute.

2. A workspace

Once logged in, create a workspace. A workspace is your isolated environment for deployments, agents, knowledge bases, and API keys. Give it a name and you're good to go.

3. Credits in your wallet

Navigate to Dashboard → Wallet and add credits. How billing works depends on the type of model you want to deploy:

Open-source models on dedicated GPUs are billed hourly from the moment the deployment becomes active — regardless of actual usage. Stop or remove the deployment when you're done to avoid unnecessary charges.

Commercial models via BYOK or Managed Keys (e.g. OpenAI, Anthropic) are billed per token — you only pay when you actually send requests.

Make sure you have sufficient credits for the model type you plan to run.

Navigate to the Model Marketplace

In your workspace, click Models Marketplace in the left sidebar. Here you'll find several hundred up-to-date models to choose from. You can:

• Search for a specific model by name using the search bar
• Filter by category — text generation, vision, embeddings, code, and more
• Load any vLLM-compatible Hugging Face model if the one you're looking for isn't listed yet

If a model you need isn't available, reach out via mycelis.ai/feedback or at info@mycelis.ai — we'll add support for it.

For this guide, we'll deploy an open-source model to a dedicated GPU. Select the model you want and click Add Model to open the Deployment Wizard.

The Deployment Wizard

The wizard guides you through four steps. Here's what each one does.

Step 1 — Scaling & Schedule

Configure how the model should scale and how long it should run each day.

User scaling: Mycelis workspaces support team collaboration with multiple members. Based on your expected concurrency, Mycelis automatically selects the most suitable GPU for you. A tutorial on team workspaces is available here: [Team Workspaces — coming soon].

Daily schedule: Define a time window during which the deployment should be active, for example 08:00–22:00. If you leave this blank, the deployment runs continuously until you stop it or remove it from your workspace.

Step 2 — Optional: LoRA Fine-Tuning

This step is optional. If you'd like to fine-tune the model on your own data before deploying it, provide a link to a JSONL dataset.

• Mycelis starts a LoRA training job before the deployment goes live
• A more powerful GPU is temporarily allocated for training
• Once training is complete, your fine-tuned variant is saved under My Models in your workspace
• The deployment then uses your customized model

If you don't need fine-tuning right now, simply skip this step.

Step 3 — Optional: OpenWebUI

Mycelis hosts a private OpenWebUI instance for you and your team. You can link your deployment here to make the model accessible directly in a chat interface.

A few things to know:

• To link a deployment to OpenWebUI, you first need to enable it under Workspace Settings → OpenWebUI
• You can add or remove the OpenWebUI link at any time, even after the deployment is already running
• To keep costs low, the OpenWebUI instance only runs when at least one deployment or agent is linked to it

Skip this step if you don't need a chat interface right now — you can always add it later.

Step 4 — Review & Deploy

The final step shows a summary of your entire configuration: selected model, GPU type, daily schedule, and estimated cost. Review everything and click Create Deployment.

Depending on the model size, your deployment will be ready in 1–3 minutes. Once it's active, it appears in your workspace with a live API endpoint.

You're live!

Congratulations — your first model is deployed on Mycelis. Here's what you can do next:

• Copy the API endpoint from the deployment card and start making requests
• Browse the API Reference to see all available endpoints and parameters
• Configure an Agent on top of your deployment
• Enable OpenWebUI under Workspace Settings and link this deployment for a chat interface