My dust.tt review examines the product, workflow, pricing approach, implementation demands, risks and best-fit buyers.
By Tolu S.5 min readDust focuses on enterprise AI assistants connected to company knowledge.
My short take: Company-specific assistants are useful to me only when the answer remains tied to the knowledge and permissions behind it. A convenient chat interface can expose confidential material if connector permissions do not survive retrieval.
| Question | Answer |
|---|---|
| What is it? | enterprise AI assistants connected to company knowledge |
| Best for | knowledge-work teams building governed assistants across internal data and tools |
| Less suitable for | organisations unable to grant secure access to the source systems agents require |
| Pricing | Sales-led unless stated otherwise below |
| Review access | Public-evidence first look; no authenticated workspace |
| Main buying test | Prove accurate, governed outcomes on representative work |
The product is designed around five buyer jobs:
Company-specific assistants are useful to me only when the answer remains tied to the knowledge and permissions behind it.
The product material in the source record describes the following areas. I would evaluate them as a connected workflow, not assume that a named feature proves an operational result.
Company-specific assistants are useful to me only when the answer remains tied to the knowledge and permissions behind it. A convenient chat interface can expose confidential material if connector permissions do not survive retrieval.
A convenient chat interface can expose confidential material if connector permissions do not survive retrieval.
Pricing evidence was recorded on 20 July 2026. The figures and plan descriptions below are historical research, not a newly verified quote. Confirm current terms with Dust before budgeting.
Dust offers packaged and enterprise plans, but current seat, usage and model allowances should be confirmed on its pricing page before procurement.
Dust is best suited to knowledge-work teams building governed assistants across internal data and tools.
It is less suitable for organisations unable to grant secure access to the source systems agents require.
I would create one assistant for a limited internal knowledge task using non-sensitive documents and two test users with different permissions. Reviewers would ask answerable, unanswerable and conflicting questions, inspect source support and test deletion and access revocation. I would measure unsupported answers and time spent checking them, then add a second source only after the first permission boundary had held.
| Alternative | Consider it when |
|---|---|
| An established category incumbent | Procurement favours maturity and a broader ecosystem |
| A specialist point solution | One narrower capability matters more than platform breadth |
| An internal build | Control and proprietary workflow are strategic |
| A services-led provider | Expert execution matters more than software ownership |
| The current process | Volume and expected benefit do not justify migration |
Before adding Dust to a shortlist, I would write down the job, the current alternative and the evidence needed to justify a change. Our guide to a requirements-first approach to choosing AI tools.
For Dust, I would also count the existing tools, review time and operating work that remain after adoption. Our guide to a framework for reviewing overlapping software costs.
I would use the AI Tool Chooser to organise the shortlist around requirements before a vendor discussion. It does not replace the product-specific, technical or professional validation described for Dust.
For related operations buying decisions, I would also read our Automata review and Beglaubigt.de review. These are adjacent evaluations, not a claim that every product replaces Dust.
I would put Dust on a conditional shortlist for knowledge-work teams building governed assistants across internal data and tools.
A convenient chat interface can expose confidential material if connector permissions do not survive retrieval.
Dust is enterprise AI assistants connected to company knowledge. This review focuses on practical buyer fit rather than repeating vendor claims.
Dust offers packaged and enterprise plans, but current seat, usage and model allowances should be confirmed on its pricing page before procurement.
Knowledge-work teams building governed assistants across internal data and tools.
No. It is a research-based first look using official public evidence. No authenticated workspace or production integration was tested.
No commercial relationship was disclosed for this review, and no rating was assigned.
Operations Lead, AI Search & GEO
Leads operations for AI search visibility and generative engine optimization (GEO) campaigns.
