It's a kind of safety net — knowing that if push comes to shove, you can fully export your entire thought collection, your second brain, in some form.

Just that feeling, that knowledge that it's , gives you the confidence to store as much as you can in a tool like this. Because you know you'd have the freedom to do something else with it if you wanted to.

So really, it all comes down to the safety net aspect.

Thank you all for sharing your feedback on this. Escalating this one with the team. We are aware of the image issue and something we will work towards. In terms of your use case for chatting with the Recall from Claude, stay tuned.

I want to highlight a specific aspect of this issue that I think deserves urgent attention beyond general data portability. The comments on exporting with attachments/possible migration are very valid, but there's an immediately usability problem.

Recall is great in many ways but it needs to work a bit better when exporting context to other AI's which is part of many people's workflows. At the minute the context files are stripped of usable images.

Images stored in Recall cards are hosted on Firebase with authenticated URLs. When a card is exported as markdown, these images appear as links rather than embedded files. This creates two serious problems:

First, those authenticated Firebase URLs will eventually expire or become inaccessible, meaning the markdown export is not just incomplete today — it will actively degrade over time, silently losing image references without any warning to the user.

Second, for those of us using Recall as a knowledge base to provide context to external AI tools like Claude, the markdown export is currently unusable for any card containing images. The only workable workaround is to copy card content into Word and save as PDF — a cumbersome manual process that shouldn't be necessary.

A full export including images saved as actual files alongside the markdown text is not a nice-to-have — it is the minimum standard for a tool that asks users to trust it with their long term knowledge. Until this is resolved, Recall cannot be fully recommended as a permanent knowledge store despite its otherwise excellent features.

I would welcome a response from the Recall team on their roadmap and timeline for addressing this. Apologies if I have double posted on this issue, an earlier attempt didn't work as far as I can see.

I want to highlight a specific aspect of the portability and knowledge re-use issue that I think deserves urgent attention beyond general data portability. I mention here the fact that exported markdown doesn't include images. Big fail.

I'm finding Recall to be an excellent tool for organising my knowledge base but we need a system that interacts with other AIs with regular exchanges to the knowledge base.

There's a big problem at present and the workarounds are clunky. Let me explain.

Images stored in Recall cards are hosted on Firebase with authenticated URLs. When a card is exported as markdown, these images appear as links rather than embedded files. This creates two serious problems:

First, those authenticated Firebase URLs will eventually expire or become inaccessible, meaning the markdown export is not just incomplete today — it will actively degrade over time, silently losing image references without any warning to the user.

Second, for those of us using Recall as a knowledge base to provide context to external AI tools like Claude, the markdown export is currently unusable for any card containing images. The only workable workaround is to copy card content into Word and save as PDF — a cumbersome manual process that shouldn't be necessary.

A full export including images saved as actual files alongside the markdown text is not a nice-to-have — it is the minimum standard for a tool that asks users to trust it with their long term knowledge. Until this is resolved, Recall cannot be fully recommended as a permanent knowledge store despite its otherwise excellent features.

I would welcome a response from the Recall team on their roadmap and timeline for addressing this.

Kind Regards and please give this some thought.

Reg

Fully agree with this. Recall's value comes notably from all the AI features, but if we need to move our knowedge base in case any issues arise with Recall, it is key to export all content. This would not make people move away from Recall, but instead reassure users

This is the most essential must-have capability. Please make this a top priority. Users should be able to enjoy data migration freedom. The exported data should include all metadata and content:

The textual portion and metadata should be an .md markdown file, with images and PDF attachments saved as is.