Watch workflows being built, step by step
Ten short animated lessons: from your first workflow to memory, MCP tools, audio and Python code.
All lessons
- 1Beginner3 min
Your first workflow: questions and answers
Three nodes, two wires: an input, a language model and an output. The base of every workflow.
- 2Beginner4 min
Text templates: an email writer
Combine several inputs into one complete prompt with {{variables}}.
- 3Beginner4 min
A translator with a fixed language
Pin the model’s behaviour with a system prompt and a locked input, so users only send the text.
- 4Intermediate5 min
Chaining two models: writer and editor
One model’s output becomes the next one’s input: a quick draft, then a careful edit.
- 5Intermediate5 min
Parallel branches: a social media pack
One text, three jobs at once: title, summary and hashtags. Independent nodes run in parallel.
- 6Intermediate5 min
Reading a receipt from a photo
Feed an image to a vision model and get the receipt’s items back as JSON.
- 7Intermediate5 min
Audio news: from text to voice
A model rewrites the text as a short news script and a text-to-speech node reads it out.
- 8Intermediate6 min
An illustrated PDF report
A model writes the report, an image model makes its cover picture, and the PDF tool puts both into one downloadable file.
- 9Advanced6 min
Answers from documentation with MCP
An MCP tool searches up-to-date docs and the model answers from those results.
- 10Advanced7 min
A chatbot with memory
Memory nodes store the conversation and recall it on the next message; each session has its own memory.
- 11Advanced8 min
Solving it with Python: an invoice calculator
Language models get exact maths wrong. Give the calculation to a Python code node and the writing to the model.
- 12Advanced10 min
Video studio: character, scene and voice
You give the character, scene, dialogue, voice and style; a model plays the director and writes a precise shot script, the video model makes the clip with synced sound, and in parallel you get a character poster and a separate voice-over.
- 13Intermediate7 min
An automatic Telegram channel: a post every day
The “Schedule” node runs the workflow every day at 9; a model writes today’s post, an image model makes its picture, and a Telegram bot posts both to the channel.
- 14Intermediate6 min
An Instagram post from one idea
You give an idea; a model writes a catchy caption with hashtags, an image model makes a square picture and the Instagram node publishes it on your page.
- 15Advanced10 min
An automatic YouTube channel: a weekly video + Telegram announcement
Every Saturday: a model writes the video’s idea, title and description as JSON, the video model makes the clip, the YouTube node uploads it and its link is announced on the Telegram channel.