August 21, 2026 · Updated October 23, 2026
The best Workflowy alternatives in 2026 (and who each is for)
Workflowy is no longer a bare-bones bullet list. It now has a Calendar, Google Calendar sync, tables, email capture, an API and integrations, AI Nodes, and Desktop MCP. The useful reasons to compare a Workflowy alternative are its plan, structure, and way of connecting to AI.
listlis is the fit for a fast, keyboard-driven tree with no monthly node quota and a hosted MCP connection. Logseq puts backlinks and queries at the center. Tana is the serious option for records, fields, and live views. Dynalist remains the closest match to Workflowy’s familiar feel, Dotflowy is the open-source copy of the Workflowy model you can self-host, SpacebarX keeps the Workflowy model on your own device, while Roam is the premium choice for people who organize their thinking as linked blocks.
Why people still look beyond Workflowy
Workflowy still does the thing that made it popular: one zoomable, nested outline that stays fast while it grows. The newer features extend that model instead of replacing it. Its Calendar gives dated work and daily notes a home; Google Calendar can sync into it. Tables add a light way to organize rows and columns. You can email notes into an outline, use its API and integrations, or work with AI through AI Nodes and Desktop MCP.
The Basic plan is free, with a 100-new-node monthly quota and 100 MB of uploads. Pro removes those limits for $6.99 a month when billed annually or $8.99 month to month. That is enough reason for a heavy outliner user to compare options. So is a preference for a graph, structured data, local files, or a remote AI connection that does not depend on a desktop app running.
Workflowy is a good choice when its outline, planning tools, and hosted sync already fit. The alternatives below make different tradeoffs.
The alternatives, and who each one fits
listlis: a keyboard-first tree with hosted MCP
listlis keeps the tree as the main interface. Each node has a title and an optional Markdown body, so an outline can stay compact until a line needs more room. Todo completion carries through a subtree and rolls up when the final child is done. Tasks, notes, and projects can live in the same shape instead of being split into separate tools.
There is no monthly node quota. listlis also takes a different approach to AI connections: its MCP server is a hosted endpoint with its own OAuth flow. A compatible client connects after a listlis sign-in, without a desktop app running. Workflowy’s Desktop MCP is useful for local work, but it needs the signed-in desktop app running.
The listlis tools understand a tree rather than a pile of disconnected notes. They can return a node with its ancestor path, siblings, and subtree; search results include their breadcrumb; and an agent can create or move a whole subtree. That fits people who want an assistant to work with a project in context while the outline remains their daily workspace.
listlis fits when the enduring shape of your work is a keyboard-driven tree and a remote OAuth MCP endpoint matters. Workflowy fits when its Calendar, tables, integrations, and exact bullet model are already the right fit.
Logseq: for networked, PKM-style thinking
Logseq is for people whose notes form a web: backlinks, daily journals, block references, and queries over a graph. It is open source and local-first, which makes it appealing if control of local data matters as much as the writing experience.
Its database version is still in beta in 2026, while the established Markdown-based app remains a separate path. Moving to Logseq includes a choice between the mature file workflow and testing the newer database approach.
Logseq fits when linked references and graph queries are part of the job. A simpler outliner fits when capture, nesting, and finding the next task are the job.
Dynalist: the closest Workflowy feel
Dynalist is the closest match to Workflowy: fast, keyboard-driven nested bullets with a familiar two-pane layout. It was built by the people who later made Obsidian.
The service, help center, and community are still online. Its own major-feature changelog, though, ends in June 2020. Dynalist can be a comfortable, settled home for a stable outline. People who want a product with a recent public feature record may prefer another option.
Dynalist fits when familiar outline muscle memory matters more than new capabilities. A newer product fits when your next few years of requirements are still expanding.
Dotflowy: the Workflowy model, open source
Dotflowy is a new browser outliner that deliberately works like Workflowy: zoom into any bullet, mirror a bullet into several places, daily pages, tags, and OPML import from Workflowy. The code is open source under the O’Saasy license, and you can self-host it on Cloudflare.
The hosted plan is cheap. Free covers 10,000 bullets, and nothing locks at the limit. Unlimited is $5 a month or $48 a year and adds a hosted MCP server and a CLI. The tradeoff is youth: it ships several releases a week, there is no App Store app, and outlines cannot be shared or collaborated on yet.
Dotflowy fits when you want Workflowy’s model, open code, and the option to run it yourself. It is the honest pick for that person over listlis, which is hosted only.
SpacebarX: the Workflowy model, stored on your device
SpacebarX is a keyboard-first browser outliner with zoom, mirrors, a Today view, and OPML import from Workflowy. Its outline lives in your browser’s storage and works offline. Optional sync goes through your own Google Drive or Dropbox rather than a SpacebarX server. Shortcuts can be remapped.
The free plan has no monthly item cap. Pro, from $4.99 a month in USD, adds Board, Table, and Mind Map views, saved searches, version history, encrypted documents, and AI with your own key. It runs in the browser or as a PWA; native desktop and mobile apps are listed as coming soon, and we found no sharing or MCP on its site.
SpacebarX fits when you want the Workflowy model with local storage and offline work. For that person it is a better pick than listlis, which is hosted and needs a connection.
Tana: structured notes without a paywall on the basics
Tana turns nodes into structured records with supertags, fields, and live searches. It also layers AI and views on top. For people who work with structured records, it is a personal database that happens to look like an outliner.
Tana’s Free plan includes supertags, live searches, and 500 monthly AI credits, so it can test the core model. Paid Plus and Pro plans begin at $8 and $14 a month, respectively, for people who need more AI, integrations, and workflow features. The real cost is the time to learn a more structured system and decide what each kind of note should be.
Tana fits when fields, supertags, and structured queries will earn their place in your work. If that sounds like paperwork around ordinary notes and tasks, an outline-first tool will feel better.
Roam Research: networked thought at a premium price
Roam made bidirectional links and the daily-notes habit familiar to a much wider audience. It is still built around interlinked blocks, so people who think by following connections can build a deep working system in it.
The price remains a real filter. Roam Pro is $15 a month, with a 31-day trial rather than a free tier. The annual option is cheaper, but this is still a paid commitment before you know whether its graph-first workflow fits.
Roam suits people who want daily notes and linked blocks to drive most of their work. For a plain outline or a tighter budget, it is more system than you need.
A quick honest comparison
| Tool | Best for | Practical tradeoff |
|---|---|---|
| Workflowy | A hosted outline with Calendar, tables, integrations, and local Desktop MCP | Basic has a 100-new-node monthly quota; Desktop MCP runs through the app on your computer |
| listlis | Keyboard-fast outlining, tasks and notes in one tree, and hosted OAuth MCP | Its focus is the tree and agent context, not a full calendar or broad integration directory |
| Logseq | Backlinks, queries, local-first PKM | Its database version remains in beta, alongside the established Markdown path |
| Dynalist | The closest feel to Workflowy | The public major-feature changelog ends in June 2020 |
| Dotflowy | The Workflowy model, open source and self-hostable, with $5 hosted MCP | Young and changing fast; no native phone app or sharing yet |
| SpacebarX | The Workflowy model with local storage, offline work, and no item cap | Browser and PWA only for now; no sharing or MCP found |
| Tana | Structured data, supertags, live searches, and AI | Free includes the core model and 500 monthly AI credits; learning the system takes time |
| Roam | Networked thought and daily notes | $15 a month after its trial, with no free tier |
The best choice matches the work you do most often. Current limits and connection options can break a close tie.
What it feels like in listlis
If your real work is speed and one place for everything, listlis keeps the loop close to the keyboard. o adds a sibling below the current line. Tab nests it under the row above, while Shift-Tab pulls it back out. A title stays short until a line needs its Markdown body. ! turns a node into a todo, and c completes it. Completing the last child marks the parent done, all the way to the root.
You can move a branch as one structural action, publish a focused node as a read-only link, or search titles and bodies. When you connect an AI client, it sees that structure too: where a node lives, what surrounds it, and what belongs below it. That is the same bet the one-tree post makes. Projects, tasks, notes, and half-formed ideas usually need to refer to one another.
Bring an OPML or Markdown outline from the import panel (here is what transfers from Workflowy), or start with an empty root. Open a root at app.listlis.com and start typing.
Still weighing options? The comparison hub lays out where each outliner fits, side by side. If speed at the keyboard decides it, the best keyboard-first outliners ranks the field by how far you get without the mouse; Dynalist vs Workflowy and Checkvist vs listlis go head to head, and MCP outliners compared covers which ones an AI assistant can actually read.