Dust
Shared AI agents over your company's knowledge, tools and models
Overview
Dust is a Paris-born platform for building AI agents grounded in a company's own data and tools, connecting Slack, Notion, Google Drive, GitHub, Salesforce and MCP servers to more than 20 models. Teams build, share and schedule agents in shared workspaces with permission controls. Pricing is per seat, from 30 a month (24 billed yearly) in euros or US dollars.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Business plan features
- Prompted to upgrade when credits run out
- 20+ models: GPT, Claude, Gemini, Mistral, DeepSeek
- Custom agents, multi-agent workflows on schedules and triggers
- Up to 3 connectors, SSO, US or EU data residency
- Everything in Pro
- For deep research and tool-heavy workflows
- Unlimited connectors and MCP servers
- Workspace-pooled credits and volume pricing
- SCIM, audit logs, single-tenant deployment, SLA
Use cases
What you can produce with Dust
- An agent that answers staff questions from Notion, Google Drive, Slack and other connected sources
- Agents shared across a workspace that any team member can invoke and improve
- Multi-agent workflows that run on a schedule or fire on an event trigger
- External tools wired into agents through native or remote MCP servers
- Permission controls separating what each agent can access from who can use it
- Conversations with agents driven programmatically through the developer API
Dust is one of the more complete platforms for putting AI agents on top of a company's own knowledge: broad connectors, MCP support, scheduled and multi-agent workflows, and a choice of more than 20 models with none locked behind a higher plan. Its permission model, which separates what an agent can reach from who may use it, is a real answer to the governance question IT will ask.
Getting value takes deliberate setup, the self-serve plan caps connectors at three and stops at 100 people, and SCIM and audit logs need Enterprise. Data residency is US or EU only. A strong choice for teams serious about running AI internally.
What we found
Dust is a platform from Paris for building AI agents that work over a company's own knowledge and tools. Stanislas Polu, a former OpenAI engineer, and Gabriel Hubert, previously at Alan, began it as a public experiment in 2022: the code has sat on GitHub under the MIT licence since August that year, and the company was incorporated in February 2023. Dust now calls its approach multiplayer AI, meaning shared workspaces where people and agents draw on the same company context, with agents built on more than 20 models from OpenAI, Anthropic, Google, Mistral and DeepSeek.
The work is in connecting data and designing agents. Dust links to Slack, Notion, Google Drive, GitHub, Salesforce, Zendesk and other systems, plus any tool exposed through the Model Context Protocol, and agents can run on schedules or triggers and hand work to one another. Governance is the pitch to IT: permissions separate what an agent can reach from who may use it. The self-serve plan includes SSO, while SCIM provisioning, audit logs and single-tenant deployment are Enterprise features. Data is hosted in the US or the EU; no Asian region is offered.
Pricing is per seat on a self-serve Business plan for teams of up to 100. A free seat carries 500 credits for the life of the account. Pro costs 30 a month, or 24 a month billed yearly, with 8,000 credits; Max costs 150, or 120 billed yearly, with 40,000 credits. The page shows euros or US dollars depending on where it is opened, at the same figures, before VAT. Credits reset monthly, bigger models and tool-heavy tasks use more of them, and the comparison table lists up to three connectors on Business. Enterprise is quoted by sales.
Dust announced a $40 million Series B in May 2026 led by Abstract and Sequoia, with Snowflake Ventures and Datadog joining, and says total funding now exceeds $60 million. It fits companies with an admin ready to own the rollout and teams that want governed agents.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Dust's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Dust unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Dust directly →
Spotted something out of date? Suggest an update →
More in Productivity
