# Dust lets every team build its own AI agents. Is that the same as an AI teammate like Maito? And what does each cost per person? Source: https://maito.app/compare/maito-vs-dust Date: 2026-10-14 Author: Dmitry Dubovetzky Maito vs Dust: Dust is a governed platform where teams build, share, and run agents across 20+ models, priced per seat. Maito is a single, ready AI teammate for your whole company in Slack and its AI assistants, for code and business work, at $149 per workspace per month plus your AI provider's usage. ## Is Dust the same as Maito? **No. Dust is a platform where teams build, share, and run many specialized agents, with an admin model for company AI and pricing per seat. Maito is one ready AI teammate the whole company works with, for code and business work.** Both work in Slack, use your company's data and tools, run on schedules and webhooks, and execute code. The question is whether your company wants to operate a catalog of agents or share one teammate. Maito costs $149 per workspace per month with unlimited people, plus your AI provider's usage and any cloud credits beyond the included $20. Dust Pro costs €30 per seat per month. ## What is Dust? **Dust is a platform for building agents from company knowledge, tools, models, and skills, and running them in a shared workspace and in Slack.** Teams build agents for work such as ticket triage, account research, analytics, and CRM updates, with more than 20 models and more than 70 connectors. Pods hold shared conversations, tasks, files, and artifacts, and admins control data and agent permissions, spending, and audit logs. [Dust](https://dust.tt/), [Dust Pods](https://docs.dust.tt/docs/user-documentation/pods/overview). Its Computer tool gives agents a temporary workspace to run code, analyze data, and edit spreadsheets, slides, and documents. [Dust Computer](https://docs.dust.tt/docs/user-documentation/agents/tools/computer). ## And what is Maito? **Maito is one AI teammate your company names and shares, ready to work once your tools are connected, in Slack and in Claude, ChatGPT, Cursor, Codex, Grok, and Muse.** Nobody designs an agent for each job. Anyone in an invited channel asks, the team joins the thread, and good results become skills and Loops. Coding tasks run in their own sandboxes with your checks and a pull request. [How Maito works](/homepage). ## How do Maito and Dust compare at a glance? **Dust gives teams the tools to build and govern many agents. Maito gives the company one teammate that's ready for business work and product changes, at one workspace price.** | | Maito | Dust | | ----------------- | ----------------------------------------------------------------- | ------------------------------------------------------ | | What it is | An AI teammate for your company | A platform for building and running agents | | Where you use it | Slack, Claude, ChatGPT, Cursor, Codex, Grok, and Muse | The Dust workspace, Chrome, and Slack | | Setup | Yes: One teammate, ready once your tools are connected | Partly: Teams design and maintain agents | | Working together | Yes: The whole team works with one agent, in any thread | Yes: Shared Pods, agents, and skills | | Business tasks | Yes: Research, reports, follow-ups, and work in your connected tools | Yes: Agents for support, sales, operations, and analytics | | Code changes | Yes: Sandbox, checks, browser QA, and a pull request | Partly: Runs code; GitHub tools create issues and review PRs | | Company knowledge | Yes: Facts and skills learned from shared work | Partly: Shared data and skills; Agent Memory is per user | | Recurring work | Yes: Loops on a schedule, from a webhook, or on request | Yes: Schedules and webhooks | | Price | $149 per workspace, unlimited people, plus provider usage | Pro €30 per seat, Max €150 per seat | ## Do you have to build agents with Dust? **Yes. In Dust, a team builds an agent for a job by choosing its knowledge, tools, model, and skills, and someone maintains it. In Maito, one teammate takes the work as people ask for it.** That's Dust's strength for a company that wants many specialized agents under central governance. Reviewers like how easily agents are built and how many models they can use, and some ask for easier ways to find past conversations. [Dust on G2](https://www.g2.com/products/dust-dust/reviews). ## Does Dust's Slack bot work the same way as Maito? **Not exactly. Dust's Slack bot, its Slack data connection, and its Slack action tools are three separate things to set up. Maito joins Slack as one teammate that anyone in an invited channel can mention.** In Dust, people mention an agent in a channel or a DM, and a channel can have a default agent or reply automatically. Connecting Slack as a data source doesn't install the bot. [Dust in Slack](https://docs.dust.tt/docs/user-documentation/agents/integrations/dust-in-slack/index). ## What do Maito and Dust do with a business task? **Maito does business tasks in Slack with your tools and what it knows about your company. In Dust, the agent a team built for that job does it.** _Illustrative conversation:_ > **Tariq:** @Maito how many tickets this month were about the mobile app login, and are they getting faster or slower to resolve? > > **Maito:** 212 tickets, up from 140 last month. Median resolution went from 3 hours to 9, mostly because 80 of them waited for an engineer. > Most are the same error after the 4.2 update. Want me to look at the login code next? ## And what do Maito and Dust do when a customer reports a bug? **Dust agents can run code, create GitHub issues, and review pull requests. Maito takes the bug from the report through a fix in your repository, with tests, QA, a pull request, and the customer reply.** _Illustrative conversation:_ > **Tariq:** Yes, please. > > **Maito:** The 4.2 app sends the login token before the clock sync finishes, so phones with a skewed clock get rejected. Recording from the emulator attached. > PR #577 waits for the sync and adds a test with a five-minute skew. Checks pass. > I drafted a reply for the 212 tickets with a workaround until the app update ships. [Dust GitHub tools](https://docs.dust.tt/docs/user-documentation/agents/tools/github). ## Do Maito and Dust both remember your company? **Dust shares company data, Pod conversations, and skills, while its Agent Memory tool is private to each user and agent. Maito learns company facts from shared threads for everyone and keeps personal memory for each person.** [Dust Agent Memory](https://docs.dust.tt/docs/user-documentation/agents/tools/agent-memory). Maito remembers things like a pricing rule or how a feature behaves, and uses them in later conversations, code and business alike. Private conversations don't teach the company unless someone asks to share a fact. [Maito FAQ](/faq). ## What do Maito and Dust cost? **Maito costs $149 per workspace per month with unlimited people, plus your AI provider's usage and any cloud credits beyond the included $20. Dust Pro costs €30 per seat per month, or €24 billed annually, so ten Pro seats cost €240 to €300 a month before tax.** **Maito: $149 per month + provider usage** - One workspace, unlimited people - $20 in included credits for coding sandboxes and web search - Your AI provider bills model usage separately **Dust Pro: €30 per seat per month** - 8,000 credits per seat each month - Monthly credits don't roll over - Max: €150 per seat for 40,000 credits Dust lets you mix seat types, so an admin can sit on a free seat and heavy users on Max. Overage can be turned on with a cap. [Dust pricing](https://dust.tt/home/pricing), [Dust seats](https://docs.dust.tt/docs/user-documentation/admins/usage-seats-and-credits/seat-management). ## When is Dust enough? **Dust is enough when your teams want to build and govern many specialized agents themselves, and code work stops at analysis, issues, and reviews.** ## So should you choose Maito or Dust? **Choose Maito for one ready AI teammate that handles business work and product changes, at one workspace price. Choose Dust if your teams want to build many specialized agents under central governance.** **Choose Maito if:** - You'd rather share one teammate than run a catalog of agents - Product bugs should end in a tested pull request - Company knowledge should be shared across the team - One workspace price, however many people ask **Choose Dust if:** - Each team wants its own specialized agents - You have people to design and maintain them - You want 20+ models and detailed governance Take one support question that turns into a product fix. Run it with a Dust agent and with Maito, and count the steps and people each one needed.