Dust

Dust is an enterprise platform for building and deploying custom AI agents that are grounded in a company's own knowledge. Agents connect to 100+ data sources and tools, use a choice of models, and can answer questions, draft content, and run multi-step tasks with memory and governance. It targets teams that want shared, reusable, governed agents rather than personal chatbots, and in 2026 uses a credit-based pricing model on top of per-seat plans.

Dust (dust.tt) is a platform that lets teams create and run custom AI agents grounded in their own company knowledge. Agents can be connected to more than 100 data sources and tools — from Notion, Slack, and Google Drive to internal systems — and combined with a choice of underlying models, memory, and feedback loops so they can answer questions, draft work, and execute multi-step tasks within an organization's context. The emphasis is on making AI "multiplayer" inside the enterprise: shared agents that teams build, reuse, and govern, rather than one-off personal chatbots. Dust provides enterprise-grade controls around permissions and data governance, and reports adoption across thousands of organizations with high weekly active usage. Dust is a venture-backed company that raised a $40M Series B in 2026 (led by Sequoia and Abstract), bringing total funding to roughly $60M+. In 2026 it shifted from flat per-seat unlimited messaging to a credit-based model, where credits are consumed based on the model and tools an agent uses — a change buyers should factor into cost planning.