Tana aims to erase the boundary between an ordinary meeting transcription service and a traditional note-taking app. Beyond recording conversations, it tries to extract structured records such as decisions, tasks, people, and projects from them. Because these records are stored in a connected knowledge graph, subsequent meetings and AI queries can draw on earlier context. The product is aimed particularly at teams with a high volume of meetings; for someone who only wants to take quick personal notes, its system may be more extensive than necessary.
Supertags and the structured knowledge approach
One of Tana’s distinguishing components is its supertag approach, which adds types and fields to records. When a line is tagged as a “task,” it can gain properties such as an assignee, date, or project; the same information can then be queried in different views. This allows an ordinary item captured in a meeting to appear in the project view and an individual’s task list without being copied again.
This flexibility does not make things easy on the first day. Establishing a sound structure across tags, fields, queries, and workspaces takes time. If every team member creates different tags before a system is established, the result may be disorder rather than the promised consistency of information. Defining a small shared vocabulary of types and structuring only frequently recurring records is more sustainable.
Moving from meetings to action
The current Tana experience brings meeting hosting, real-time transcription, and the creation of documents, tasks, and decisions from a conversation into the same workflow. It also offers an agent that it says can record Zoom, Microsoft Teams, and Google Meet sessions without placing a visible bot in the meeting. “Bot-free” does not mean that the meeting is not being recorded or that participants do not need to be informed. Organizational policy, local law, and participant consent should be considered before recording begins.
Presenting AI-generated results as suggestions that await user approval, rather than allowing them to change the database directly, is an important design choice. It allows an incorrect assignee, invented deadline, or out-of-context decision to be spotted before becoming permanent. However, approving large numbers of suggestions in bulk without considering them brings back the same data-quality problem. In critical meetings, the transcript should be compared with the decision text, and ambiguities should be corrected by the meeting owner.
The Today view brings together meetings, pinned work, and a daily capture area. AI chat can answer questions using context from the workspace. MCP and various integrations make it possible to create workflows between information in Tana and external tools. Although this capability is powerful for technical teams, they should examine which areas can be read and written before granting broad connection permissions.
Plans and the true cost
According to the official pricing, the free plan includes five hosted meetings with AI transcription per month, one calendar connection, and 50 AI queries. Content can still be viewed and edited on the free tier. Pro’s early price is $20 per user per month, while the page lists a regular price of $30. This plan includes approximately 20 times the AI usage of the free version, unlimited hosted meetings, the bot-free agent for external meetings, integrations, and MCP. Max is listed at an early price of $80 per month and a regular price of $120, with five times Pro’s AI capacity and additional support. Since extra credits can be purchased after the AI quota is exhausted, heavy users should also account for costs beyond the subscription.
Final verdict
Tana is a strong candidate for teams that want meetings to become institutional memory and trackable work, rather than merely being archived. Product managers, consultants, and executives with back-to-back conversations may see genuine value in linking decisions to projects and people. The free allowance of five meetings is reasonable for testing transcription accuracy and the suggestion workflow with real conversations. By contrast, the Pro fee is difficult to justify for individuals who rarely attend meetings, those who only want a task list, or teams unwilling to spend time building a structure. A purchase decision should be based not on an impressive AI demo, but on the number of tasks accurately captured and genuinely completed during a one-month trial.
Research sources: https://tana.inc/pricing https://tana.inc/learn https://tana.inc/blog/how-ai-is-changing-knowledge-management-2026