---
title: Deploy from Chat
url: "/docs/ai-agents/first-blossom-chat/"
canonical_url: https://docs.blossom88.com/docs/ai-agents/first-blossom-chat/
markdown_url: https://docs.blossom88.com/docs/ai-agents/first-blossom-chat.md
summary: Preview and deploy a FastAPI App in a conversation with Blossom.
categories:
- learn
- ai
---

# Deploy from Chat

Ask your AI app to preview the FastAPI Quickstart, deploy it, and hand you the
live URL. Start a chat with Blossom connected; to connect ChatGPT, Claude, or
another MCP client, follow [Connect AI Apps](/docs/ai-agents/connect-ai-apps/).



<aside class="docs-callout not-typography" data-tone="info">
  
  <svg xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24" stroke="currentColor" aria-hidden="true"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M13 16h-1v-4h-1m1-4h.01M21 12a9 9 0 11-18 0 9 9 0 0118 0z"></path></svg>
  
  <div class="docs-callout-body">
    
    <div class="info-content">
      
<p>You can preview your App before setting up billing. When you’re ready for your
first Blossom Cloud deployment, <a href="/docs/billing/">activate Cloud</a>
once for your Organization.</p>

    </div>
  </div>
</aside>



## 1. Preview FastAPI


> **Ask your AI agent:** Which Blossom Account am I using, and what access do I have? Preview the
> FastAPI Quickstart on Blossom Cloud. Show me the App name, Location,
> Environment, and web Service. Stop at the preview.



The preview shows what would run before anything starts:

| Preview | FastAPI Quickstart |
|---|---|
| App | `python-fastapi` |
| Location | Falkenstein (fsn1) · Blossom Cloud |
| Environment | `production` |
| web Service | `uvicorn main:app --host 0.0.0.0 --port $PORT`, port 3000 |
| Size | Micro · 0.25 vCPU / 437.5 MB, 1 Container |

Ask for another name or Location and preview again. **Read only** access is
enough to preview. To deploy, connect with **Full access**.

## 2. Deploy


> **Ask your AI agent:** Deploy the plan we just reviewed and follow it to completion. Give me the
> live App URL and a link to its Environment in Blossom. If the deployment
> fails, show me the first error in the logs.



The first deploy takes a few minutes while Blossom builds the example. Your AI
app follows it and replies with both links.

## 3. Open the App

Open the Environment link and select **Activity**. The deployment is labeled
with the AI app that started it, and **View App** opens the live example.

![A succeeded deployment labeled via Codex via MCP, with Capacity, Build, Deploy, and Available checked and a View App button.](/assets/images/getting-started/chat/activity-light.png)


![The FastAPI example App open in a browser at its onblsm.dev URL.](/assets/images/getting-started/web-ui/live-app.png)


## Try your own App


> **Ask your AI agent:** Preview my Git repository on Blossom. Explain which Services it would run,
> how each one starts, and what it needs before deploying. Stop at the preview.



To compare costs before deploying, tell it what you pay today and ask for a
monthly estimate for the same workload and region.

## What You Get

- A `python-fastapi` App, live on its own HTTPS URL
- The deployment's steps and logs under the Environment's **Activity**
- A chat that can preview, deploy, and check on your Apps