July 31st, 2025

You can now choose the LLM model you want to use for chat. Currently, our default is to use โClaude Sonnet 4โ.
July 29th, 2025

The preview page now supports a browser console. Click โConsoleโ on the bottom left to see your appโs logs, warnings, and errors.
July 25th, 2025

Hercules can now search the web to find images to add to your app! Just ask chat to find images or let our agent find them itself
July 25th, 2025

Hercules is now available to everyone! Head to hercules.app and start building!
July 24th, 2025

Hercules Chat now streams the code for file edits directly inline. It also shows the number of lines of code added/removed. The code streaming UI will collapse once the edit is finished.
July 24th, 2025

Hercules now supports multiple chat sessions per app. Click the + button to start an empty chat and click the the history icon to switch between your chats.
We strongly recommend creating new chats after about 10-20 messages for more accurate results
July 24th, 2025

Ever forget to send something in your prompt, donโt want to wait for it to finish, but donโt want to cancel it and restart? Queued messages is your answer.
Send a message while an existing prompt is running and Hercules will queue it, and then automatically run it as soon as your existing message finishes.
Full screen example

July 24th, 2025

Are we even a venture backed company if we donโt support dark mode? ๐
Change your theme by clicking your user icon in the top right and choosing system, light, or dark
July 23rd, 2025

Hercules now letโs you see previous versions of your app.
Hit the dropdown button next to the publish button to rewind time โฐ
July 22nd, 2025

It now takes ~7 seconds for you appโs development to spin up, down from ~30 seconds. We will continue to push to make this instant!