What's the difference between Free, Pro, and Max seats?

These are seat types in the Business plan. Free is for occasional users (500 lifetime credits). Pro ($24/mo) has 8k credits/mo for most users. Max ($120/mo) has 40k credits/mo for power users.

Which AI models are available on Dust?

All plans include 20+ models from OpenAI (GPT), Anthropic (Claude), Google (Gemini), Mistral, and DeepSeek. You choose the model per agent; no model is locked to a higher plan.

How does the dual-layer permission model work?

It separates what data/tools agents can access from which users can invoke those agents. This prevents privilege escalation and allows granular, auditable access control.

Can I build custom agents with my own tools?

Yes. Agents can use custom skills, knowledge sources, and tools. You can connect data via 70+ connectors or build custom MCP servers for proprietary systems.

Does unused credit roll over to the next month?

No. Each seat's monthly credit allocation resets at the start of every billing period to keep pricing predictable.

What happens when I run out of credits?

An in-progress response will finish. After that, Pro/Max users can continue through workspace overage if enabled by an admin. Free users are prompted to upgrade.

Directory/Automation/AI Assistants/Dust

Dust

Dust is a platform for building and running AI agents that collaborate with humans across company workflows. It connects to 70+ data sources like Slack, Notion, and CRM systems, giving agents deep context to perform real work. Users can choose from 20+ frontier and open-source models (GPT, Claude, Gemini, etc.) for different tasks. The core capability is creating custom AI agents with specific skills, knowledge, and tools. These agents can handle complex, multi-step tasks like incident response, lead qualification, content drafting, or data analysis. It emphasizes 'multiplayer AI,' meaning agents and human teams work together in shared workspaces (Pods) with real-time collaboration features. Dust is built for enterprises, offering strong security with SOC 2 Type II, SSO, SCIM, data residency (US/EU), and granular permission controls. It targets teams across engineering, sales, support, and marketing who need to automate repetitive tasks while maintaining oversight and control.

Automation

Visit website

Added on

Sep 22, 2026

Monthly Visits

455K

Multi-agent orchestration20+ LLMs70+ connectorsSOC 2 Type IIMCP serversShared workspacesSSO & SCIM

Product information

Everything worth knowing about Dust

The complete picture — from everyday features to the technical detail builders and IT folks dig for.

Core features

What it actually does

Agents with Company Context

Connect Dust to 70+ connectors including Slack, Notion, Google Drive, GitHub, Salesforce, and Zendesk. Give agents the context they need to get work done.

Multiplayer AI for Cross-functional Work

Bring people and agents into the same workflow. Dust helps teams coordinate complex work across departments using shared context, reusable skills, connected tools, and human review.

Model Flexibility

Build agents with 20+ frontier and open-source models from OpenAI, Anthropic, Google, Mistral, and more. Choose the model that fits each workflow.

AI-native Governance

Control how AI is used across your company. Set permissions for data, tools, systems, and agents, then use audit logs and analytics to monitor activity, adoption, usage, costs, and ROI.

Technical capabilities

Dual-Layer Permission Model

Separates what agents can access from who can use them. SCIM-synced groups, admin-gated overrides, zero privilege escalation, and audit logs let teams move quickly while keeping company data locked down.

Enterprise Deployment Options

Supports US or EU data residency, dedicated single-tenant deployment, and custom data retention policies for enhanced security and compliance.

Must watch videos

Dust for GTM engineers | Dust online event

34:40

Tutorial34:40

Dust for GTM engineers | Dust online event

Your sales team has a list of event leads, but they’re waiting on you to clean it up, enrich the contacts, and prepare the outreach. What if they could do it themselves using workflows you’ve built? Join Maxence, go-to-market engineer at Dust, for an online event showing how to turn a messy contact list into enriched, segmented leads ready for Lemlist campaigns. See how reusable skills and shared company context help sales teams handle the process, with a human review before anything gets sent. Explore Dust: https://dust.com

Build your GTM cockpit | Dust online event

29:58

Tutorial29:58

Build your GTM cockpit | Dust online event

The meeting’s over, but the work isn’t. You still need to write the follow-up email, update your team, and schedule the next call. Join Elia from Dust’s GTM team, hosted by Romain, for an online event showing how to turn Granola meeting notes into Gmail drafts, Slack updates, and Google Calendar invites. Watch the workflow in action using a reusable skill in Dust, with human review before sending emails or creating invites. Explore Dust: https://dust.com

The Governance of AI | Dust online event

36:46

Tutorial36:46

The Governance of AI | Dust online event

Your employees may already be using AI, even if your company hasn’t approved a platform. How do you move from shadow AI to a deployment your IT and security teams can confidently support? Join Zeid, security at Dust, for an online event on AI governance. Learn how to assess the risks of specific use cases, define what agents can access and do, and preserve the evidence your security team needs when something goes wrong. Explore Dust: https://dust.com

Who uses it

Real use cases, no hype

Engineering Operations

AI-powered code debugging, automated reviews, and incident response runbooks.

Customer Support

Automated ticket triage, context-aware response drafting, and escalation detection.

Sales Enablement

Instant account research, lead qualification, and proposal generation from past wins.

Marketing & Content

Campaign brief generation, content localization, and performance analysis into narratives.

Data & Analytics

Self-service analytics via natural language, pipeline monitoring, and report drafting.

Cross-Functional Workflows

Coordinate complex tasks across departments using shared context and human review.

AI Governance & Admin

Manage permissions, audit logs, and usage analytics across the entire AI stack.

What's great

Unifies company context across 70+ tools (Slack, Notion, CRM)
Model flexibility with 20+ LLMs from top providers
Enterprise-grade security (SOC 2 Type II, SSO, SCIM)
Enables complex multi-agent workflows with triggers
Supports human-in-the-loop review and coordination

Technical strengths

Dual-layer permission model for granular access control
Open-source core and developer API for extensibility
Advanced features like context compaction and structured output

Where it falls short

Complex pricing with credit-based consumption
Steep learning curve for non-technical users
Advanced features locked behind Enterprise plan
Requires significant setup to connect all company data
Credit costs can escalate with complex, tool-heavy agents

Technical limitations

Dependency on external LLM provider pricing and availability
Customization depth may overwhelm teams without AI expertise

For technical folks

The deep-dive specs

FAQ

Includes technical Q&A

A credit is the unit for measuring AI usage, consumed per message based on model choice and tool actions. More complex tasks with search or app actions use more credits.