Q: Is there blocks to insert python code?
1. Is there block to insert python code?
2. Are Meta integration the official ones (WhatsApp, Facebook and Instagram)?
3. Doest it include a chat visualizer or can I use my preferred service like chatbot?
4. Is there a limit in the size of my model or amount of data to train my agent?
5. Is there several kinds of outputs or chat only? Can I generate MPC executions in function of some agent decision, for example to generate an email, to send a sms, to add a data into a db or run an script? Of course all with my own keys.
mario_moreno_millanVerified purchaser
release0
Sep 28, 2025A: Hi mario_moreno_millan
Here’s how Release0 currently works for each of the points you raised:
1- Running Python code
Release0 doesn’t have a “Python code block” that executes raw Python inside the agent flow.
Instead, you can run any server-side logic by:
• Calling an external service or function via the HTTP Request block
• Hosting your Python code on your own server, cloud function (e.g., AWS Lambda, Google Cloud Functions), or a backend like Supabase/NocoDB
• Returning the result back to the agent in real time
This keeps the agent secure and flexible while still letting you execute anything you can write in Python.
Docs: https://docs.release0.com/editor/blocks/integrations/http-request
2- Meta integrations (WhatsApp, Facebook, Instagram)
Our current native WhatsApp integration is built on the official WhatsApp Cloud API from Meta.
• Guide: https://docs.release0.com/integrations/whatsapp/overview
For Facebook Messenger and Instagram, native integrations are not yet available. You can request them on the roadmap here: https://release0.com/feature-request
Meanwhile, you can connect to these channels today through middleware using our offical integrations with Make.com and Zapier or using the HTTP Request block.
https://docs.release0.com/editor/blocks/integrations/make-com
https://docs.release0.com/editor/blocks/integrations/zapier
3 - Chat visualizer or front-end of your choice
Release0 provides:
• An embedded chat widget for websites (HTML/JS or WordPress plugin)
• A shareable hosted chat link (e.g., https://r0.ar/youragent)
If you already use another chat front-end or a custom app (e.g., your own chatbot UI), you can bypass the built-in visualizer and just interact with the agent via our REST API or webhooks. So you’re free to keep your preferred client while Release0 handles the backend logic.
https://docs.release0.com/api-reference/authentication
4 - Model size and data limits
Release0 doesn’t retrain models itself. Instead it uses BYOK (Bring Your Own Key) with the provider you choose (OpenAI, Anthropic, Groq, DeepSeek, etc.).
• You can upload large knowledge bases (docs, PDFs, Google Docs) and use retrieval-augmented generation (RAG) so the LLM always cites and answers strictly from your data.
• There’s no hard size limit imposed by Release0 other than practical provider constraints (e.g., chunking large files for retrieval).
• You can combine many large documents and keep adding to your knowledge base.
5 - Output types and automations beyond chat
Release0 isn’t limited to chat replies.
An agent can make decisions and then trigger actions such as:
• Send an email (via Email block and your own SMTP provider)
• Send an SMS (through Twilio or similar using the HTTP Request block)
• Write to a database (Google Sheets, Supabase, NocoDB)
• Call external APIs or run a script (your own endpoint or serverless function)
• Tag users, branch logic, or chain agents
In other words, you can build a fully automated flow where the “conversation” is only one piece, and your backend executes whatever is needed.