comparison8 min read

Reentry vs Heptabase: Which workspace fits the way you think?

Heptabase helps you study and map what you deliberately want to understand. Reentry starts earlier, when you save something before knowing why it might matter.

By Dennis WeishaarUpdated July 17, 2026

The short answer is that Reentry is the stronger choice for people who want to collect first and still get research, learning, resurfacing, and action later. Drop in the link, video, PDF, screenshot, or thought without reading it, filing it, tagging it, or knowing why it matters yet. Reentry understands it now and brings it back when it can help.

Heptabase is the better fit when the study ritual itself is what you want: read deliberately, write in your own words, and manually map what you understand. Reentry can support that learning too, but it does not require the ritual before your knowledge becomes useful.

Reentry vs Heptabase at a glance

If you care most about...	Better fit	Why
A deliberate read, write, and map study ritual	Heptabase	Its workflow is designed around active learning, cards, whiteboards, highlights, and connected notes.
Saving before you know why it matters, without filing or tagging it	Reentry	Drop it now. Enrichment, labels, search, and resurfacing keep it useful even when the canvas stays messy.
Current cross-platform clients, sync, and collaboration	Heptabase	It already ships across desktop, web, and mobile with real-time collaboration.
Social and web source enrichment	Reentry	Reentry extracts reusable context, transcripts, media, engagement, audience reactions, summaries, and labels instead of merely preserving the page.
Learning from what you saved without planning a course yourself	Reentry	Tutor turns workspace context into researched mini-courses, lessons, and knowledge checks automatically.
An agent that can research and reorganize the workspace it lives in	Reentry	The agent can search, create notes and canvases, group and place material, and use connected research tools.
Proactive reminders, generated feeds, and mini-courses	Reentry	Feed and Tutor repurpose workspace context without requiring a new prompt for every useful return.

The real difference is what happens before organization

Heptabase describes itself as an intelligent visual knowledge base for students, researchers, and lifelong learners. Its product language centers on mastering difficult topics, reading sources, writing notes, and arranging ideas on whiteboards. That is a coherent job: you know what you are trying to understand, and Heptabase gives you a serious environment for doing the work.

Reentry starts earlier and keeps going.

Most things are saved before you know why they matter. You notice a useful Reddit thread during lunch, send yourself a screenshot, keep a two-hour video for later, or leave 47 tabs open because closing them feels like losing something. At that moment, you do not want to design a taxonomy or write a durable note. You want to keep the thing without creating another job.

That is Reentry's core promise: drop it once and let Reentry do the remembering.

You can organize a Reentry canvas carefully when spatial structure helps. You can also let it become chaotic. Library, search, labels, related context, Feed, Tutor, and the agent continue working underneath it. You do not have to earn future usefulness by maintaining a perfect map.

Capture and source understanding

Heptabase supports notes, PDFs, highlights, YouTube transcripts, journals, whiteboards, and other learning-oriented objects. Its current site specifically presents PDF, YouTube, note, and journal context as sources its AI can use. For deliberate study, that is already a broad and useful foundation.

Reentry puts more emphasis on the transition from an external source into durable context.

When you drop a source, Reentry preserves the original, extracts its text and metadata, stores full transcripts, analyzes social images and carousels, reads engagement and audience discussion, summarizes what matters, and creates context the rest of the workspace can use. A social post does not become a card whose most meaningful content is a cookie notice or login screen.

This is especially relevant for people who collect from YouTube, Reddit, TikTok, Instagram, X, blogs, product pages, PDFs, and other sources. Reentry's model is different from embedding the page and leaving interpretation for later: the saved object becomes usable context when it enters the workspace.

Visual organization

Heptabase is effective at making deliberate structure feel tangible. Cards, whiteboards, mind maps, sections, backlinks, tags, tables, and highlights support a workflow where the user reads, writes in their own words, and maps relationships. If that structured study ritual is the activity you specifically want, Heptabase is purpose-built for it.

Reentry's canvas is also spatial, but the canvas is not the whole product. It is the place where sources, notes, groups, drawings, and agent actions remain visible together. The agent can create and manipulate that structure, while Feed, Library, Graph, Tutor, and search provide other ways back into the same material.

The distinction is subtle but important:

In Heptabase, mapping is a central part of how you deepen understanding.
In Reentry, mapping is one useful representation of a memory system that should still help when the map is untidy.

AI chat and agent capabilities

It would be inaccurate to describe modern Heptabase as "just chat with your notes."

Heptabase now documents both AI Chat and AI Agent modes. With Space search enabled, it can search across a Space, select relevant cards and whiteboards, and send a bounded subset to the model. Its MCP integration can search objects, inspect whiteboards, retrieve full objects, search PDFs, create note cards, and append to a journal. Reentry also ships an MCP server, so MCP access is shared rather than a Heptabase advantage.

Reentry's agent is designed as a resident operator inside the workspace. It can use saved context, search web and social platforms, inspect posts and comments, create Markdown notes, create and organize canvases and groups, move or duplicate placements, schedule local research, use packaged and user-provided skills, and focus the result for the user.

The difference is not "learning versus workspace action." Reentry does both.

Reentry uses AI for learning, research, retrieval, and writing back into the workspace. Tutor turns workspace context into researched mini-courses. Feed proactively creates useful briefings and brings older material back into view. Search and chat retrieve saved sources and full transcripts, while the agent can research beyond the collection, create Markdown notes, and write the result back where the work is happening.

What Reentry adds is that this intelligence is also proactive and operational. It can surface a relevant source without requiring the user to remember or name it, extend saved material with fresh research, and then place, group, label, or organize the result. Heptabase supports a deliberate read, write, and map workflow. Reentry supports learning and research while also letting the workspace resurface, teach, and act.

Proactive resurfacing

A blank chat still depends on remembering what to ask.

Reentry adds surfaces where old context can return without a perfectly phrased request:

Feed creates short, automatically curated entries from active and forgotten workspace material.
Tutor generates guided mini-courses from the workspace and can extend the material with researched context.
Related context and search reconnect an item with other placements, notes, and sources.
The agent can bring an older source into a current question even when the user did not name it directly.

This is the largest conceptual difference between the products. Heptabase assumes meaningful labor before and after AI: read, write, connect, and build understanding. Reentry is for the material you saved before you knew whether you would ever perform that labor.

Local access, cloud sync, and availability

Heptabase is currently distributed across web, desktop, and mobile capture, with offline access, real-time sync, and collaboration. Its current plans also provide different AI credit and model allowances, with bring-your-own-key options.

Reentry is currently a private Mac beta with an iOS capture wrapper. Its core canvases, notes, Library, search data, chats, jobs, transcripts, enrichments, and cached assets are stored locally. Choose Heptabase when current team collaboration or its broader client distribution is the deciding requirement.

Choose Heptabase if...

You are deliberately studying a subject and want to read, annotate, write, and map it.
Cross-platform access, collaboration, and cloud sync are requirements.
You want to annotate sources manually and turn them into linked study cards and whiteboards.
You enjoy maintaining structured cards, tags, whiteboards, and connections.

Choose Reentry if...

You save far more than you revisit.
Your inputs are a messy mixture of links, social posts, videos, screenshots, PDFs, and notes.
You want to save without reading, filing, tagging, or deciding what something is for.
You want useful context extracted when you save something, not only when you ask about it later.
You want research and generated learning without manually turning every source into study material.
You want forgotten material to return through more than search and chat.
You want an agent that can research, write, and reorganize the workspace itself.
You prefer a local-first Mac workspace with an iOS capture wrapper.

Which one should you use?

Heptabase helps you work on knowledge you have chosen. Reentry is being built for the knowledge you almost lost before you knew to choose it.

Sources and current product references

Private beta

Drop it once. Let Reentry do the remembering.