Turn a request into code
Let AI understand the project, edit files, run checks, and leave a result you can keep working on.
CaoGen is an AI workspace that helps you finish the work. Start with a goal. Finish code, research, content, and everyday work.
State the goal. CaoGen organizes the AI, tools, and steps, then keeps the work moving toward a result you can inspect and deliver.
Open source · Current public release v0.1.8 · macOS Intel signed, notarized, and stapled
What you can do
From one request to a deliverable result, CaoGen is built for work that takes more than one step.
Let AI understand the project, edit files, run checks, and leave a result you can keep working on.
Gather context, pull out what matters, and shape a report that helps you decide what comes next.
Start with a topic, plan the script and assets, and move step by step toward delivery.
Work with files, sheets, and documents through a clear process instead of scattered tasks.
Start with a goal
Give CaoGen the work. It helps plan, create, check, and package the result instead of returning one chat reply.
Read structure, entry points, and constraints.
Put risks and priority into the task.
Change code in an isolated workspace.
Collect inspectable execution evidence.
Check diffs, risks, and open items.
You decide whether to merge, repair, or stop.
Use AI like a team
You do not need to move between separate windows. CaoGen keeps understanding, creating, checking, and packaging together.
Shape a one-line goal into clear steps and next actions.
Create code, documents, reports, designs, and media.
Pause, change direction, or take over without losing progress.
Keep working in one place
CaoGen keeps the goal, process, files, and result in one desktop workspace. Start in chat, or open the project and control room directly.
Inspectable and extensible under AGPL-3.0-only.
Keep projects and runtime data close to your machine by default.
Bring the providers and keys you trust.
Start with one thing
Download CaoGen, write one goal, and start with the first step.
Describe the result first. Choose the AI and tools after that.