The first release of Launch Week #3 is the Morph framework update.

The Morph Framework

Morph is a full-stack framework that enables you to build AI and data apps using Python (Requires Python 3.9+ and Node.js 18+).

Docs: https://docs.morph-data.io/docs/ja/getting-started/why-morph

This update enhances features specifically related to AI app development.

💬 stream_chat() + <Chat /> : AI App Toolkit

We’ve added a Python function called stream_chat() and a <Chat /> component as part of our toolkit for constructing AI apps.

With these tools, you can build custom LLM chat interfaces or multi-agent chat systems in Python and then directly convert them into web applications.

Check out the demo video to see how you can build an AI chat app using Langchain in just ninety seconds:

🪄 Deploy with morph deploy

Once your app is ready, you can deploy it to the cloud so your team members can see it.

Simply run the morph deploy command in your terminal to deploy.

Currently, deployment is supported only via the command line, but support for CI/CD via GitHub integration is planned for the near future.

For documentation on deployment, check here:

The Ideal Scenario for Using Morph: Langchain → Web App

The ideal scenario for using Morph is to prototype AI workflows and multi-agent processes in Python, and then seamlessly convert them into web applications.

For example, imagine your team has an idea for a custom AI application. To test its feasibility, you might first experiment with Python and the command line using LLM workflow frameworks like Langchain or multi-agent frameworks like CrewAI.

Once the workflow logic is sufficiently developed, how do you share it with your team? Do you build an application from scratch using Django or Flask? While Streamlit or Gradio might be good options, do you have the deployment environment set up?

In moments like these, remember Morph. With just a few lines added to your Python code, Morph builds the application server for you, and with simple Markdown edits, it constructs the frontend as well!

More Advanced Customization

Another key feature of the Morph framework is its high customizability. Since it is a Python framework, you are free to add and use any Python packages. Additionally, the following options are available:

Edit the Dockerfile to Add OS-Level Packages

When you initialize a project with morph new, a Dockerfile is generated. By editing the Dockerfile, you can add OS-level packages during deployment.

Customize the Frontend

Morph’s page development uses MDX files. With MDX, you can use React components—meaning you can develop your own React components or add component libraries via npm packages.