GPTree meets you where you already are: your ChatGPT history, the tabs you think in, the tools your team runs on, the API keys you already pay for, even your local machine. Nothing starts over.
Your data stays yours. Import what you choose. Export anytime.
Start from your existing conversations or capture new ones mid-flow.
One upload brings every ChatGPT conversation in as a searchable tree. Batch upload handles big archives; nothing is skipped silently.
The Chrome extension adds a branch action inside chatgpt.com, lifting a live thread into GPTree with context intact.
Keep working where you are. When you need branching, open GPTree from the chat and continue without restarting your thinking.
A connected tool is not a shortcut menu; it is context. Your inbox, calendar, and trackers feed real state into conversations and routines, so answers stop being generic and briefings stop being manual.
"What needs me today?" pulls from your real calendar and inbox, on a branch, without you pasting anything.
Sending, posting, and editing always wait for your explicit approval, with an approve-with-edits option to fix drafts inline.
Per-user, encrypted at rest, revocable in one click. Disconnecting revokes access immediately.
Two ways to power the same workspace. Most people mix them.
Zero setup: GPTree runs the models and one meter covers everything. Every plan includes a monthly allowance (2,000 on Free, 10,000 on Pro, 25,000 per user on Team), premium models draw more per request, and top-up packs from $5 never expire.
Best for: getting started, everyday questions, and never thinking about provider accounts.
Every planConnect your own OpenAI, Anthropic, Google, Mistral, Azure, or other keys. Usage bills to your provider accounts at provider rates, keys are encrypted at rest, and removal is one click. Your key, their bill, our branches.
Best for: heavy usage, teams with existing provider contracts, and cost control at volume.
Pro and TeamMixing is normal: keep credits for quick questions and run heavy branches on your keys. Per-branch model choice works the same either way.
A pip-installed Python client runs code and reads files on your machine while every tool call lands on a branch you can compare, replay, or merge. Outbound connection only; nothing on your machine listens.
Once you merge the winning branch, export clean output to your tooling.
Ship it into docs fast. Clean, portable format that works everywhere.
Headings, bullets, decision logs structured for immediate use.
Publish merged branches as pages. Sync your best outputs automatically.
Coming SoonExport merged branches as notes. Build your second brain with AI insights.
Coming SoonChatGPT history export
Import
Chrome extension
Capture + Branch
GPTree app
In-flow branching
Notion export/sync
Two-way sync
Obsidian export
Export to vault
Google Docs / Word / PDF
Document import
Start free in under a minute. No card required. Use built-in credits or bring your own keys.
