Import from Obsidian and Notion
Migrate an existing vault or export into Orcha with hierarchy, relations, attachments, and links preserved.
Orcha imports Obsidian vaults and Notion exports as ZIP archives, through a guided flow in the app under Migrations. Imports preserve folder hierarchy, relations, attachments, and links.
Prepare the archive
- Obsidian: zip your vault folder.
- Notion: use Notion's export (Markdown and CSV), which downloads as a ZIP.
Uploads are validated defensively: archives with path traversal, symlinks, oversized expansion, suspicious compression ratios, or executable attachment types are rejected.
Preview first
Every import starts with a dry run. Upload the archive, pick the destination workspace, and choose a conflict policy:
| Policy | Behavior when a matching item already exists |
|---|---|
update | Update the existing item (default) |
skip | Leave the existing item untouched |
duplicate | Import as a new copy |
The preview report shows what would be created, updated, and skipped, and checks your plan allowances before anything is written: total storage, indexed content, the archive size, and the largest single attachment in the archive. Free workspaces cap one attachment at 25 MB and the archive at 500 MB; paid plans allow 1 GB per attachment and a 5 GB archive. An archive that trips a limit names the offending file in the preview and cannot be started, so a large asset fails before the import rather than halfway through it.
Run, monitor, resume
Start the run from the preview. Progress is visible on the Migrations page, and a run that fails or is cancelled can be resumed; reruns use persisted source IDs, hashes, and destination hashes, so the default update policy never overwrites a destination you edited after import.
Roll back
Within seven days, a run can be rolled back. Rollback removes imported objects that have not been touched since the import and leaves anything you edited in place. The full fidelity and execution report is downloadable per run.
After the import
Imported files index automatically and become searchable through retrieval. Imported binary attachments are stored as assets and served through authorized asset URLs. Image embeds render inline; embeds of other attachment types (PDFs, spreadsheets, audio) render as download links to the stored asset.