Now live: Audio Transcription 
→ 
The new “Transcribe Audio” block lets your agents pull transcripts from any audio URL — using GPT-4o, GPT-4o-mini, or Whisper.
Need users to upload their own files? Use the new “Upload Audio” user input and set Processing = Transcribe.
Plus, 10 new blocks for transforming text 
Everything you actually need to do with text — now built in.
No prompt required. Just drop the block in and go.
Here’s what’s new:
Summarize Text – Get the TL;DR
Improve Writing – Fix spelling, grammar, clarity
Translate Text – Any language, instantly
Rewrite for Audience – Adapt for different personas
Convert Reading Level – Make it simpler or more advanced
Merge Text – Combine multiple inputs into one
Adjust Tone – Casual, formal, fun — your call
Align to Brand – Stay on voice, every time
Rewrite Text – Give it a rule, and it’ll rewrite at scale
Convert to Format – Repurpose content into a blog, email, post — whatever you need
All of this is available now in the menu block.