TODO for AI
vs 
Manus
By Marcell Havlik, founder of TODO for AI · Updated
A general-purpose agent that researches and builds documents, slides, sites and videos in its cloud, with a desktop app that can work in approved local folders. Here is where each one wins, with numbers, and who should pick which.
| Fact | Manus | TODO for AI |
|---|---|---|
| Interface | Web, desktop (macOS, Windows), mobile | Web, desktop, mobile, CLI, IDE (Zed, JetBrains, VS Code), voice |
| Models | Vendor-selected, or your own key via Manus Flex | Any provider: Claude, GPT, Gemini, Grok, Ollama |
| Source | Closed | Edge + CLI public |
| Starts at | $20/mo | Free |
Key takeaways
- Models. Manus model choice: Vendor-selected, or your own key via Manus Flex. 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. Manus pricing: Credit based plans from $20 for 4,000 credits plus 300 daily refresh credits. Heavy tasks burn credits fast. 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. Manus gets one task done mostly in their own cloud. 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 Manus if you want a polished consumer agent for research, slides, sites or video. Pick TODO for AI if you want a Manus alternative that runs on your own Linux, macOS or Windows machine, plus an always-on cloud VM included in the plan.
Manus
TODO for AI
Any providerWe are not competing for the same job
Manus gets the task done.
TODO for AI organises the whole company’s work.
Manus answers one question: how do I get this task done, mostly in their own cloud? 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 Manus is, and where it stops
Manus popularised the general-purpose cloud agent: give it a goal, it browses, researches and hands back a document, a deck, a site or, since Manus 2.0 in September 2026, a first-cut video. It needs no setup. For a non-technical user who wants a finished artefact from one prompt, it is a strong choice.
It is no longer cloud-only. The desktop app's My Computer feature works in folders you authorise and runs command-line tools, with approvals you can grant once or always, and Browser Operator drives your local Chrome. Manus Flex (September 2026) lets you bring your own API key through partner providers such as OpenRouter. What remains: credit pricing that is hard to predict, local access that goes through a desktop app on macOS or Windows, and projects where colleagues share instructions and files but not each other's tasks.
People who want polished research, decks, sites or short videos from one prompt, mostly in a vendor cloud.
Three reasons Manus users look elsewhere
Cost per task is hard to predict and heavy work is expensive. Flex moves inference to your own provider bill, but the credit plans remain the default.
Local work goes through the Manus desktop app on macOS or Windows, in folders you add one at a time. There is no Linux machine or server of your own to run it on, and the always-on option is a paid Cloud Computer.
By default Manus routes between vendors for you. Flex adds bring-your-own-key through its partner providers; there is no cross-vendor review of one model's work by another.
A project holds shared instructions and files and you can invite colleagues to it. Tasks are private by default and shared one at a time, so the project shows what everyone agreed to publish rather than what everyone is working on.
Six dimensions that decide it
Manus 12/30TODO for AI 29/30How does Manus compare to TODO for AI, feature by feature?
| Feature | Manus | TODO for AI |
|---|---|---|
Persistent cloud VM An always-on machine. Files, logins and sessions survive between runs. | Cloud Computer (paid add-on) | Yes |
Model choice | Vendor pick, or BYOK via Flex | 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. | Browser Operator extension | Yes |
Business tasks, not only code SEO, email, ads, social, CRM via CLI tools. | Yes | Yes |
Any CLI as an integration | Local CLI via desktop app | 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. | No | Yes |
Many agents, each configured Each agent gets its own model, prompt, tools and permissions. Pick which one takes an item. | One agent | Unlimited, per task |
Projects for the whole company Engineering, marketing, ops, finance: separate projects and agents, one team and one bill. | Shared, tasks opt-in | Unlimited, shared |
Team roles, pooled usage | Per seat | Roles + pooled |
Tool permissions | Basic | Allow, approve, block |
Open runtime | No | Edge + CLI public |
Same ladder, different meter
Abstract credits, about 200 per dollar, plus a 300-credit daily refresh. Plan credits reset monthly; add-on credits never expire. A heavy task can burn a plan in an afternoon. With Flex, model inference bills to your own provider.
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.