Your files shouldn’t decide which AI model you can use. Now they don’t. Upload once, switch between any model, and keep the conversation going without paying to parse the same file again.
We've improved file parsing across all models! Attach any file to any chat model.
No more switching models first. Babbily now reads PDFs, Word, Excel, PowerPoint, code, images, audio, video, and ZIPs on every chat model... and reads each file once, so follow-ups don't pay for it again.
Finishing up a major update to file parsing.
Our goal is to simplify and unify the model picker so users don’t have to know what every model can or can’t do.
Instead, Babbily will handle those capability gaps behind the scenes, giving models access to capabilities they don’t natively support and letting you focus on choosing the model you want to use.
Speaking of our new character glow-ups: Fun style, way more personality.
Soon you'll get to build your own Babbily avatar into one of these looks, for more personalization.
Today’s Babbily release is about making repeatable work easier: organize it in Projects, automate it with Scheduled Tasks, and bring in app context through Plugins.
If you used Connectors before, reconnect your apps in Settings > Plugins. Credentials weren’t transferred.
Now live in Babbily ✨
📁 Projects keep chats, instructions, files, and memory together.
⏰ Scheduled Tasks automate recurring prompts on paid plans.
🔌 Connectors are now Plugins, with a new Composio-powered connection flow.
Existing apps may need to be reconnected.
...Fable 5.1 will be added shortly.
Small workflow upgrades matter.
Babbily now supports optional completion notifications, so users do not have to keep checking whether a chat, image, or video is finished.
New in Babbily: push notifications.
Start a chat, image, or video generation, switch tabs, and we’ll let you know when it’s done.
Less waiting around. More getting things made.
Manage it anytime in Settings > Account > Notifications.
Small upgrade, big quality-of-life win: Babbily now supports passkeys on web.
You can sign in with Face ID, Touch ID, Windows Hello, or a security key after adding one in Settings > Account. Faster for users, better for account security.
Passkeys are now live on Babbily.
Add Face ID, Touch ID, Windows Hello, a password manager, or a security key to sign in faster and more securely.
Available now on web in Settings > Account.
Excited to see Babbily Finance and our updated News experience go live.
This release helps users move from information to understanding faster, with markets, company pages, news, citations, and follow-up chat all in one place.
Babbily Finance is here, alongside a refreshed News experience.
Users can now track markets, explore company pages, read source-backed financial stories, review prediction markets, and move from breaking news to deeper AI-powered context in one workspace.
News and Finance are now connected the way people actually think: what happened, why it matters, and what to ask next.
From an ops perspective, Canvas Mode is a huge unlock for Babbily users.
Create, edit, export, render JSON, and push work to connectors without bouncing between tools.
Even better: edits made in Canvas stay in context, so the model keeps building from the latest version.
Canvas Mode and JSON Reneering is now on Babbily.
Now you can create, edit, export, and send AI-generated work without leaving the chat.
With Canvas, you can:
- Edit documents, tables, code, and structured outputs directly in Babbily
- Render JSON into cleaner visual components
- Download work as Word, PDF, CSV, Excel-ready files, or JSON
- Push finished work directly to connected apps
- Keep your Canvas edits in context for follow-up prompts
That last part is the big unlock: when you edit inside Canvas, Babbily knows what changed. Your next prompt continues from the current version, not the old chat message.
Babbily is becoming more than a place to talk to AI. It’s becoming a place to actually work with it.
AI is going to make a lot more people feel capable.
More people will build things, learn faster, solve problems, and create work they are proud of.
That is the part I am most excited about.
One thing we’ve learned building Babbily: AI products need great docs, not just great features.
As Babbily becomes a true AI workspace, we needed a better way to help people actually use it.
New docs are live:
https://t.co/6Khhvkfwid
We just launched the new Babbily Docs site. This is really about making AI something people can actually use. Not just talk about. Not just wonder what model to pick or how files, memory, tools, and connectors are supposed to work.
Start here:
https://t.co/lN37OFnPhs
We just launched the new Babbily Docs site. This is really about making AI something people can actually use. Not just talk about. Not just wonder what model to pick or how files, memory, tools, and connectors are supposed to work.
Start here:
https://t.co/lN37OFnPhs
We're playing catch up since rebuilding Babbily. Things we're adding in short order:
- Context windows
- Context Summarization (will include when switching models mid thread)
- Memory moving out of beta
- Skills (and skill creation)
- Slash Commands
- Connectors
- Auto mode
Anthropic really said "what if cron jobs were smart" and shipped Claude Code Routines. schedule it, connect your repo, close your laptop, go touch grass. it just runs.