TODO for AI
vs 
Linear
By Marcell Havlik, founder of TODO for AI · Updated
The issue tracker engineering teams like, with its own Linear Agent and coding sessions in a Linear sandbox. Here is where each one wins, with numbers, and who should pick which.
| Fact | Linear | TODO for AI |
|---|---|---|
| Interface | Web, desktop, mobile | Web, desktop, mobile, CLI, IDE (Zed, JetBrains, VS Code), voice |
| Models | Linear Agent: Claude, GPT or GLM, picked per workspace; delegated agents bring their own | Any provider: Claude, GPT, Gemini, Grok, Ollama |
| Source | Closed | Edge + CLI public |
| Starts at | Free | Free |
Key takeaways
- Models. Linear model choice: Linear Agent: Claude, GPT or GLM, picked per workspace; delegated agents bring their own. TODO for AI runs any provider, Claude, GPT, Gemini, Grok or a local Ollama model, on one plan, and one model can review another's work.
- Benchmark. TODO for AI scored 89.9% on Terminal-Bench 2.1 with Opus 5.5 (high) (own harness, 80/89 tasks, 2026-10-08; not an official leaderboard submission).
- Price. Linear pricing: Free, Basic $10, Business $16 per user per month billed yearly, Enterprise custom. Coding sessions draw AI credits: model tokens at provider rates plus $0.25 per 20 minutes of sandbox. TODO for AI has a free plan (Claude Sonnet only), then Starter $20, Pro $100 or Ultra $200 a month with any provider included.
- Where the work happens. Linear gets one task done inside the tracker and its sandbox. TODO for AI is one board your team and agents work through together, on your own PC, a cloud VM and your logged-in Chrome.
- Who should pick which. Pick Linear if your work is engineering issues and cycles, and you want the best tracker for them. Pick TODO for AI if the work on your board is not only code.
Linear
Linear Agent: Claude, GPT or GLM, picked per workspace; delegated agents bring their own
TODO for AI
Any providerWe are not competing for the same job
Linear gets the task done.
TODO for AI organises the whole company’s work.
Linear answers one question: how do I get this task done, inside the tracker and its sandbox? So do the others — Cursor inside the editor, Claude Code in one person's session, Lindy by asking its assistant.
We answer a different one: where does the work live? One board per project, and a project for each part of the company — engineering, marketing, ops, finance, one per client. Your colleagues work on it. So do your agents, as many as you want to run, each with its own model, tools and permissions. Items carry the people who worked them and the agent that ran them, and the board filters by person, so what the company has in flight is one screen instead of a Slack thread.
That is not a team plan bolted onto an AI tool. It is the product. The agents are how items get closed; the board is what the company runs on.
What Linear is, and where it stops
Linear is the closest thing to what we are building, from the opposite direction. It already has the thing the agent vendors do not: a real shared board, where a teammate and an agent can hold the same item, where state is queued, started, blocked and done, and where everyone can see it. It now has its own worker too. Linear Agent answers in chat, in comments and in Slack, creates and updates issues, runs Skills and scheduled Loops, reaches other tools over MCP, and starts coding sessions in a Linear-hosted sandbox that installs the toolchain, tests the app in a browser and opens the PR. Admins pick Claude, GPT or GLM for the workspace.
The shape is still product and engineering. The sandbox works on a GitHub repo, the browser inside it is for testing your app, and anything else goes through whatever MCP connections exist. There is no shell on your own machines, no session in your logged-in Chrome, and the model is one workspace setting rather than a choice per agent.
Engineering teams who want issues, cycles and coding agents on one board.
Three reasons Linear users look elsewhere
Linear Agent can reach other tools over MCP, but the sandbox, the browser and the workflow are shaped around a repo. Invoices, outreach and ad accounts have no natural home there.
Coding sessions run in Linear's managed sandbox. Linear never gets a shell on your box or a session in your Chrome, so a tool behind your logins without a connector goes back to a human.
Admins choose the model for coding sessions across the workspace. TODO for AI sets the model per agent, so triage can run on a cheap model and the hard part on a strong one.
Six dimensions that decide it
Linear 17/30TODO for AI 29/30How does Linear compare to TODO for AI, feature by feature?
| Feature | Linear | TODO for AI |
|---|---|---|
Persistent cloud VM An always-on machine. Files, logins and sessions survive between runs. | Linear sandbox for coding sessions | Yes |
Model choice | Claude, GPT, GLM per workspace | Any provider |
Cross-vendor review Review GPT output with Claude, or the other way around. | No | Yes |
Drives your real browser Browser extension (Chrome, Firefox) plus cloud browsers. | Sandbox browser for testing, not your session | Yes |
Business tasks, not only code SEO, email, ads, social, CRM via CLI tools. | Engineering first, MCP beyond | Yes |
Any CLI as an integration | MCP + Agent API | Any binary |
Voice control | No | Yes |
People and agents on one board Colleagues and agents work the same list. Items show who contributed and which agent ran them; filter the board by person. | Yes, for engineering issues | Yes |
Many agents, each configured Each agent gets its own model, prompt, tools and permissions. Pick which one takes an item. | Linear Agent + 30+ delegated | Unlimited, per task |
Projects for the whole company Engineering, marketing, ops, finance: separate projects and agents, one team and one bill. | Teams and projects | Unlimited, shared |
Team roles, pooled usage | Per seat + AI credits | Roles + pooled |
Tool permissions | OAuth scopes | Allow, approve, block |
Open runtime | No | Edge + CLI public |
Same ladder, different meter
Seats, plus AI credits for coding sessions and Loops (tokens at provider rates, sandbox $0.25 per 20 minutes). Delegated agents such as Devin or Cursor bill on their own plan on top.
One pool priced in dollars of model cost, spendable on any vendor. Session and weekly windows pace it, and a 1.3-1.5x plan multiplier applies, so a $20 plan is about $15 of list-price API. Team seat prices go into one shared pool.