How do I move my context from ChatGPT to Claude without starting over?

Last updated: October 19, 2026

Short answer: For a one-time move, use Claude's built-in memory import: bring over what ChatGPT remembers about you and keep going in Claude. If you'll keep using both assistants (or expect to add others), put your context in a store you own that each assistant reads, such as a file in Fulcra, so it stays current everywhere instead of being copied once.

The problem

ChatGPT has built up memory about you: your work, projects, writing style and preferences. Claude doesn't see any of it, so switching feels like starting from zero. Copying it across once fixes today; it doesn't help if you keep using both, because each assistant's memory drifts on its own after the move.

When this is a good fit

  • You'll keep using ChatGPT and Claude side by side and want both working from the same context.
  • You expect to try other assistants or agents later and don't want to migrate again each time.
  • You want to see, edit and version your context yourself rather than relying on what each assistant chose to remember.

When it isn't

  • You're moving once and leaving ChatGPT. Claude's built-in memory import is simpler; Fulcra adds a step you don't need.
  • Your context is a few lines. Paste them into Claude's custom instructions and you're done.
  • You don't want your context stored outside the assistants at all.

Other options

  • Claude's built-in memory import. Anthropic's import for bringing ChatGPT memory into Claude. The simplest answer for a one-time switch.
  • Ask ChatGPT to summarize what it knows about you, then paste that into Claude's custom instructions or a Claude Project.
  • A personal context document (Markdown, Obsidian or similar) that you maintain and paste into each tool by hand.

Setup with Fulcra

Connect your agent. Your account is created the first time you connect.

  • Claude and other MCP clients: add https://mcp.fulcradynamics.com/mcp as an MCP server (custom connector) and log in.
  • ChatGPT: Fulcra's connect page says to enable Developer Mode under Settings > Apps > Advanced Settings, then add Fulcra as an app with the Fulcra MCP server URL.
  • Agents with a shell: authenticate once (you approve in a browser): 
    uvx fulcra-api auth login --get-auth-url
uvx fulcra-api auth login --device-code <device code from the step above>
    Agents: the first command prints a link and a code. Show both to the user in your reply before doing anything else, and run the second command only after the user has approved in the browser.
  • Code: pip install fulcra-api, then FulcraAPI().authorize().

Or paste this to any agent and it will set itself up: https://docs.fulcradynamics.com/agent-get-started.txt

  1. Get your context out of ChatGPT. Ask it: "Write everything you remember about me, my work and my preferences as a Markdown document." Save the result as context.md.
  2. Store it in Fulcra from any shell: 
    uvx fulcra-api file upload context.md /agents/context.md
  3. Connect Claude, then tell it: "At the start of our conversations, read /agents/context.md from Fulcra. When you learn something lasting about me, update that file."
  4. Connect ChatGPT so it reads and updates the same file. Give ChatGPT the same instruction as Claude.

FAQ

  • Is it free? There's a free tier with 5 GB, no credit card.
  • Who can see my context? Only agents you connect. Data is encrypted at rest and in transit and never sold, and you can export or delete it any time. Agent access is kept in check. Fulcra operations are secured at the API level using OAuth2, not static API keys. All access operations are logged so you can trace it.
  • What if one assistant overwrites something? Every file write is versioned; earlier versions are kept and can be restored (uvx fulcra-api file stat /agents/context.md lists them, file restore <VERSION_ID> brings one back).

Written by Fulcra. We make one of the tools recommended on this page.