I didn't start testing Todoist because I was looking for a new app. I needed to survive one genuinely messy week, and my usual system of Slack reminders, notebook lists, and a calendar I trusted too much was leaking tasks. So I gave Todoist a proper run: three ongoing projects, about twenty tasks, one week, no imported data, no prior setup. Here's how it actually went.
Step one — I set up the week as projects, not piles
The first thing I did was create three projects: Client Work, Moving Errands, and Side Project. The instinct was to dump everything into one giant list, but Todoist's structure pushes you toward projects, and that turned out to be the right call. Everything went in through the quick add bar, mostly typed like normal sentences: "send revised invoice Friday 10am" or "water plants every 3 days."
The natural language parsing is what stood out first. Most dates and recurrences were understood on the first try — "every 3 days" became a proper recurring task with zero extra clicks, and "Friday 10am" just worked. One task I typed as "draft weekend plan next week" landed on the wrong Monday, and I had to drag it over manually. Fast, but not faultless.
Step two — what Todoist's views and the AI part actually did
The workflow that clicked for me was the morning routine: open the Today view, see everything due, and reorder it into a rough sequence. I also ran a filter for "p1 & today" to isolate the two or three things that genuinely had to happen. That short list saved me more than once when the day tried to pull me in six directions.
Since people keep arguing over whether this is the best AI task manager 2026 has to offer, I tested the AI side too. Todoist's AI Assist lives on the paid tier and can suggest dates, break tasks into smaller steps, and rewrite entries. It worked — technically. Most suggestions were conservative, though, like re-dating an overdue task or adding a label. I ignored half of them. If you're hoping for an AI smart to do list app that decides what deserves your attention, this isn't that. It's bookkeeping help, not judgment.
Step three — where the free plan stopped me
Here's the friction I hit, and it's worth knowing before you commit. The free tier works indefinitely, but it's capped at five active projects. I had two old projects sitting around from a failed attempt years ago, and once I added my three test projects, that was the limit — I had to archive something just to create a fourth. Reminders, labels, and the AI assistant are all behind the paid plan. If you're evaluating Todoist as a free AI to do list app 2026, the free plan is genuinely usable — but the wall shows up faster than the marketing suggests.
The tradeoff that actually matters
The real question isn't whether Todoist works. It does. The question is whether you plan in the way Todoist thinks. It's project-based through and through — everything hangs off projects, labels, filters, and due dates. That's excellent for managing recurring responsibilities against deadlines. But if you're actually looking for a daily planner that lays goals and the whole week out side by side, the way taskly does, Todoist can feel heavy, like a database wearing a to-do list's clothes.
If you don't mind maintaining your own structure and you live by deadlines, Todoist is a strong fit. If you plan more visually and want the week's shape to tell you what to do, a simpler planner approach might save you real effort. I'm still not sure I'd reach for Todoist for creative work with no deadline attached — its strengths are recurring tasks and pressure, not inspiration.
I ended the week having missed fewer deadlines than the week before, which is the only honest measure I care about. Todoist gets out of the way once you accept its structure, and its free tier is enough to test that properly. The AI features aren't the reason to pay, in my view. What matters is that it's a tool you'll actually reopen on a Tuesday morning — and Todoist cleared that bar for me.
Comments
Leave a Comment