Dust

What is Dust AI?

Dust AI is a team AI workspace for building and using agents around company knowledge and tools. Instead of repeatedly pasting the same background into a separate chat, a team can configure assistants for particular jobs. A company knowledge assistant is useful when people need to find an approved answer or prepare a draft from information the business already maintains.

Make the data connection useful

Dust AI connects supported services such as document, communication and development tools. Connected company data can reduce repeated setup, but access and freshness still matter. A shared agent platform should be organized around what each group is allowed to see. Before adding a large source, decide who owns it, which pages are current and how outdated material should be removed from the agent’s context.

A short example

A small agency could use Dust AI to prepare a client handover checklist from its approved project notes. Ask the business AI assistant to list unfinished actions and point to the supporting record. Use a fictional client for the pilot. If a delivery date is missing, the assistant should say so rather than infer one from an older project with a similar name.

Start with one agent

Create a focused assistant in Dust AI with a short job description, a small source set and a predictable output. Test five known questions and one unknown question. For workspace AI agents, a simple internal helper is easier to assess than a system that reads everything and can change several apps. Keep the first version read-only until the answers and source access are clear.

Review sources and permissions

Check whether a cited document actually supports the answer. Try an account that should not have access to a restricted source. Ask about a recently changed policy and confirm which version was used. These checks help distinguish a confident summary from a useful company answer. Do not place secret information in a broadly shared source just because an agent can technically retrieve it.

Automation needs an owner

Dust AI supports multi-agent work and scheduled agent tasks in its platform. A schedule is useful only if someone knows what should happen when the task fails or the source changes. For a recurring brief, define the recipient, the date range and the missing-data behaviour. Review a few runs manually before relying on the result in a customer conversation or operational decision.

A useful check before you continue

Keep an owner and last-reviewed date on important source collections. When a policy changes, test a question about that specific change. This helps the team distinguish a model problem from stale company material and makes it clearer who should correct the underlying information.

Budget by seat and workload

Dust AI uses seat-based AI pricing, with Free, Pro and Max seat types in the Business plan. Free credits are lifetime rather than monthly, and paid monthly allocations do not roll over. More demanding models and tool-heavy tasks can use more capacity. Programmatic usage has a separate displayed credit rate, so include automated runs as well as people’s chat activity in the budget.

Plans and pricing

Who it suits

Dust AI is worth considering for teams that need shared assistants connected to maintained business information. It may be more than a solo user needs for occasional writing. Begin with a task that has known answers, confirm source permissions and compare the correct useful outputs with the seat and usage cost. Expand the source set only when the team can still explain where answers come from.

Common questions

Are the free 500 credits renewed monthly?

No. The current Free seat allocation is lifetime.

Can a team mix seat types?

Yes. Current Business documentation allows Free, Pro and Max seats to be assigned according to individual usage.

Check the official plans and access details before choosing your setup.

For related options on SimplifyAITools, see Lindy and Dify.