Tana

For a decade, users have been trapped in a digital architecture of their own making: the folder. Whether it’s the physical filing cabinets of the 20th century or the nested hierarchies of Notion and Evernote, data has lived in silos. Tana Labs is betting everything on the idea that this is fundamentally wrong. Tana is an EverythingOS—a node-based environment where every line of text is a discrete object capable of existing in multiple places at once.

Based in Oslo, Tana Labs has emerged from a long beta period with a product that positions itself less as a tool and more as a philosophy. It sits at the intersection of structured data and fluid thought, targeting the “knowledge worker” who has outgrown the limitations of linear note-taking. Unlike its predecessors, which often feel like digital paper, Tana is designed as a relational database disguised as an outliner. It moves away from the page-centric model and towards a global knowledge graph, where the relationship between ideas is as important as the ideas themselves.

The market positioning here is aggressive. Tana isn’t looking to be your “second brain” in the way Obsidian or Roam Research might; it wants to be the central nervous system for professional workflows. By integrating high-level AI models directly into the core architecture, Tana avoids the “AI as a sidecar” trap that plagues its competitors. It doesn’t just help you write; it aims to help you organize, query, and automate the very fabric of your information. However, this level of sophistication comes with a learning curve that will likely rebuff casual users.

• Supertags: The centerpiece of the Tana design. A Supertag is a template that instantly injects a node with predefined fields, logic, and data structures. Applying #meeting to a line doesn’t just color the text—it adds fields for attendees, date, and action items.
• Compound Intelligence: Tana leverages a mix of GPT-4 and Claude to perform “agentic” tasks. It doesn’t just summarize; it can extract action items from a voice memo and automatically categorize them into project trackers based on the logic of your Supertags.
• Tana Capture: A dedicated mobile companion designed for high-friction environments. It focuses on rapid-fire input—voice-to-text, photo scans, and quick notes—which are then piped directly into the desktop’s “Inbox” for later processing.

To understand Tana, you must first accept that the page is a lie. In Tana, there are no pages—only nodes. This is a radical departure from the skeuomorphic “notebook” metaphor that has dominated the industry. The strength of this approach lies in its granularity. Because every entry is a node, the entire workspace can be queried with a level of precision that makes Notion’s databases look like toys.

The Power of the Supertag

The Supertag system is the most significant innovation in the productivity space since the introduction of the backlink. In most apps, adding a tag is a way to filter. In Tana, adding a tag is an act of definition. When you tag a node as #Person, you are telling the system that this object has a phone number, an email, and a “last contacted” date. This creates an ontology of your life. The friction of setup is real—you essentially have to build your own system from the ground up—but the payoff is a workspace that understands the type of data it holds. This enables Live Queries, where you can call up every #Task that is #HighPriority and #Overdue across your entire graph, regardless of where it was originally written.

AI as an Architect

While every developer is currently slapping an AI chatbox into their sidebar, Tana’s implementation of Compound Intelligence is deeply structural. Its core purpose is data transformation, not just text generation. The design promise is that a disorganized 20-minute brainstorm can be transcribed, with Tana identifying the distinct #Projects mentioned and creating #ActionItems for each. That is where the “EverythingOS” pitch starts to feel real. It targets the cognitive load of manual organization, which has always been the Achilles’ heel of high-power tools like Obsidian.

The Learning Wall

However, the friction is real. Tana is not an “open and play” application. The onboarding experience is a daunting gauntlet of concepts like “nodes,” “fields,” and “semantic search.” For the average professional just looking to keep a grocery list, Tana is an absurd over-correction. It demands that the user become a knowledge architect. If you don’t enjoy tweaking your workflow and defining your own data structures, Tana will feel like a chore rather than a utility.

The aesthetic direction is clinical and focused. It eschews the playful “blocks” of Notion for a more streamlined, text-heavy interface that prioritizes information density. A node-based workspace of this scope is inherently demanding, and the desktop-first architecture is built to handle large graphs in a way web-based editors often struggle with. The UI design uses a muted palette, which is necessary given how much information can be on-screen at once. It reads as a professional workstation, not a social media app.

The experience is, by design, lopsided. Tana is fundamentally a desktop-first engine. The iOS app (Tana Capture) is built as a streamlined funnel for getting data into the system, not for managing the graph. The Android version appears to be the weaker sibling, with a narrower feature set than its Apple counterpart. If a workflow relies heavily on managing complex data on an Android device, the mobile apps are strictly capture tools, and the desktop remains where the real work happens.

Tana is a glimpse into a future where our tools are as smart as the data we put into them. It is a sophisticated, uncompromising platform that demands a lot from its users but offers a level of organizational power previously available only to those who could write custom code. It isn’t for everyone, and it certainly isn’t for those who want a simple place to jot down thoughts. But for the right audience—people drowning in a sea of disconnected notes—it is a capable and forward-thinking option. We recommend it to system builders and heavy-duty knowledge workers, with the caveat that the learning curve and mobile limitations are real costs.