Dust

What is Dust?

Dust is an enterprise platform for teams to build, deploy, and manage customizable AI agents that integrate with company data and tools. It lets users create no-code agents powered by models such as GPT-4 and Claude, and connect them to sources like Google Drive, Notion, Slack, GitHub, and SharePoint. The platform also supports actions through built-in tools, custom Dust Apps, native integrations, and a developer API for programmatic control. Dust is for product teams, ops teams, and enterprises that want AI agents to work across internal knowledge and workflows.

Key Features

• Retrieval Augmented Generation (RAG): Dust retrieves relevant data from connected sources before answering, which helps reduce hallucinations and improves accuracy across documents and chats.
• Multi-tool Agents: These agents can choose and chain multiple tools during reasoning, which helps automate multi-step work without manual setup for complex queries.
• Agent Memory: Dust stores memory for each user-agent pair, so agents can remember preferences, projects, and past interactions across conversations.
• Query Tables: This feature combines semantic search with SQL-like analysis on spreadsheet and CSV data, so non-technical teams can work with tabular data more easily.
• Vision Support: Dust supports image analysis with GPT-4o, GPT-4-turbo, and Anthropic models, so workflows can include screenshots, diagrams, and other visual inputs.
• Scheduled Agents: Available on Pro, Team, and Enterprise, scheduled agents run automatically for recurring reports, summaries, or data syncs and support hands-off automation.
• Spaces: Spaces add permission layers for data access and agent visibility, which helps teams control who can see what in shared environments.
• Connections: Native integrations connect sources such as Slack, Google Drive, Notion, BigQuery, and SharePoint, so Dust can pull work data into one AI workspace.

Use Cases

• Post-Sales Operations Manager: Uses Dust agents to automate repetitive post-sales workflows across CRM, support tickets, and email systems. At Profound, this reclaimed 1,800+ hours per month and removed manual ticket processing bottlenecks.

• Content Marketing Manager: Builds specialized agents for content transformation and localization using company templates, brand guidelines, and localization rules. At Alan, customer story production dropped from 2 days to a few hours, and case study turnaround fell by 80%.

• Localization Operations Lead: Configures a localization agent with market-specific adaptation rules for emails, web pages, and social content. At Qonto, this reduced content localization time by up to 70%.

Strengths and Weaknesses

Strengths:

• G2 reviewers (rating 4.9/5, 22 reviews) give Dust a high overall score. Secondary coverage notes a limited sample size and sparse negative sentiment.
• Public review data in the research set shows sparse negative sentiment on G2. That suggests generally positive user feedback in the available sample.

Weaknesses:

• The G2 rating is based on 22 reviews, and secondary coverage notes a limited sample size. That means the public sentiment data is narrow.
• The research set does not include specific recurring complaints, support quotes, or reliability quotes. That limits how much detail we can verify about common drawbacks.

Pricing

• Pro: €29/user/month. Month-to-month plan with advanced models, custom agents with executable actions, connections for GitHub, Google Drive, Notion, and Slack, plus native integrations for Zendesk, Slack, and the Chrome Extension. Includes unlimited messages under fair use, free credits for programmatic usage with fixed-price overages, up to 1 GB/user of data sources, and one private space. A 14-day trial is available with no credit card required.
• Enterprise: Custom. Includes everything in Pro, plus multiple private spaces, larger storage and file size limits, custom pricing for programmatic usage, SSO (Okta, Entra ID, Jumpcloud), SCIM user provisioning, US/EU data hosting options, Salesforce Tool, dedicated account management, priority support, and flexible payment options. Designed for 100+ member organizations.

No perpetual free tier is listed. Dust offers a time-limited trial only, and pricing is published in EUR.

Who Is It For?

Ideal for:

• GTM or operations lead at a 20 to 500 person SaaS team: Dust fits growth to scale-up companies that keep knowledge across Slack, Notion, and Google Drive. It works best for non-technical teams that want internal assistants on top of company data.
• Sales or product thinker at a scale-up: It suits people who need research agents that pull from tickets, internal narratives, and shared docs for personalized sales or product work. The fit is strongest when they want to do this without coding.
• Internal builder, analyst, or enablement lead in a knowledge-heavy company: Dust can help with churn analysis, positioning updates, and onboarding questions through semantic search and extraction across company sources. It also fits HR or enablement teams that want broader access to internal knowledge.

Not ideal for:

• Solo founders or very small teams: Dust depends on shared workspaces, connected company data, and broad internal use, so lighter options like ChatGPT custom GPTs or Claude Projects may fit better.
• Developers who want full code control, or teams that need deep vertical tools: Dust is a horizontal product and can feel too abstract for custom agent logic or domain-specific work, so LangChain, LlamaIndex, Harvey, or Casetext are better matches depending on the use case.

Use Dust if your company has information spread across tools like Slack, Notion, Drive, Confluence, or GitHub and non-technical teams need agents on top of that data. Skip it if you are working alone, want a code-first agent framework, or need a tool built for a narrow field such as legal work.

Alternatives and Comparisons

• Glean: Dust does more than enterprise search. It combines direct answers with sources, access across 100+ tools, and agentic automation that can act on information. Glean does pure semantic search and intranet indexing better in large enterprises. Choose Dust if you want search plus no-code workflow actions, and choose Glean if retrieval depth is the main need. Switching difficulty from Glean is medium, based on the research.

• Salesforce Agentforce: Dust does cross-tool search and automation better for teams that work across Slack, Notion, Google Drive, Salesforce, and other apps, and it supports multiple AI models. Salesforce Agentforce does Salesforce-native CRM use cases better, especially for sales and service teams that want built-in governance tied to Salesforce data. Choose Dust if you need broader enterprise search and faster setup in hours, and choose Agentforce if your work is centered on Salesforce.

• Gumloop: Dust does enterprise knowledge search better, with RAG-based agents that synthesize information across tools such as Slack, Drive, and Salesforce, and public materials also cite SOC2 Type II, HIPAA, and GDPR. Gumloop does natural-language workflow building better for smaller teams that want quick setup and lower entry pricing. Choose Dust if permissions-aware knowledge synthesis is the priority, and choose Gumloop if simple workflow creation matters more than deep enterprise integrations.

Getting Started

Setup:

• Signup: Public signup requirements and trial terms are not described in the available sources.
• Time to first result: No public time estimate was found. The first screen is an empty dashboard, and initial setup starts with creating an agent, then connecting tools and data sources, and setting its capabilities.

Learning curve:

• Dust is described as no-code and intentionally designed around guided learning. It is not presented as a simple "hello world" style start, and there are no sample templates listed in the research.
• Beginner: no public estimate found. Experienced: no public estimate found.

Where to get help:

• Official help is available through Dust's support portal and email. Dust states it responds to all requests and commits to a response within 2 business days. An @help agent can answer basic questions instantly, and it does not require signup.
• Learning materials include a quickstart guide, Dust Academy, and an official tutorial video.
• Community support appears small but tight and growing. Answers come from both Dust staff and community members, while third-party learning content is still limited.

Watch out for:

• Self-service customers were reported as significantly underonboarded.
• Enterprise customers previously needed face-to-face onboarding training, which suggests setup could be harder without guided help.

Developer Experience

Dust is a no-code platform for building AI assistants and workflows through natural language and a visual interface. Based on the available research, its developer surface is aimed more at non-technical teams than at programmatic integration. Public source material provided for this section does not include documentation feedback, code examples, or reported time-to-first-result details.

What developers like:

• The available research describes Dust as accessible through natural language and a visual builder, which reduces the need for direct coding.

Common frustrations:

• The available research positions Dust primarily for non-technical teams, so developers looking for a code-first workflow may find limited programmatic depth in the surfaced materials.

Security and Privacy

• Encryption: Dust states that data is encrypted with AES-256 at rest and TLS 1.3 in transit. (security page)
• Data handling: The vendor states that customers own their data, user data is not used for training, and zero data retention is available with no data stored by third-party model providers. (security page)
• Data residency: Dust states that data residency is available in the EU and US. (security page)
• Access control: The vendor states that role-based access control, SCIM, SAML SSO, audit logs, and custom data retention duration for workspace conversations are available. (security page)
• Compliance: Dust states that it is SOC 2 Type 2 certified, GDPR compliant, and HIPAA compliant. (security page)
• Security testing: The vendor states that it runs a bug bounty program on HackerOne. (security page)

Product Momentum

• Release pace: Users describe Dust's shipping pace as steady, with multiple feature drops in recent months. The changelog is active and notes both additions and deprecations.
• Recent releases: Dust added an interactive workspace analytics dashboard and public API on April 8, 2026. Other recent releases include an Amplitude integration via MCP, and Google Drive write access with real-time builder observability.
• Growth: Public signals point to a stable to growing trajectory, and Dust is VC-backed. Its ecosystem is still expanding through integrations such as Amplitude, Granola, Intercom MCP, and Google Drive write.
• Search interest: Google Trends data in the provided period shows no clear direction, with +0.0% change between the first and second half. The latest interest score is 0/100, and the peak score is also 0/100.
• Risks: No notable controversy or trust issues were found in the provided sources. Dependency on external LLMs remains a product risk, though abandonment risk appears low given the active changelog and support outreach.

FAQ

What is Dust?

Dust is a platform for building custom AI agents that use company data to answer questions and automate work. Public documentation describes it as a no-code product for deploying agents across teams.

What is Dust used for?

Dust is used to create AI agents that connect to internal knowledge and business tools. Research indicates teams use it for search, question answering, and workflow automation.

Does Dust support RAG?

Yes. Dust uses retrieval augmented generation, or RAG, to pull relevant information from connected sources before generating a response.

Can Dust connect to company data?

Yes. Research describes Dust as built around connected organizational data, with examples that include sources such as Slack, Notion, and Google Drive.

Is Dust no-code?

Yes. Public materials describe Dust as a no-code platform for building and deploying custom AI agents.

Is Dust free?

No perpetual free tier is listed in the research. Dust offers a time-limited trial, and paid plans start at €29 per user per month for Pro.

What does the Pro plan include?

The Pro plan is listed at €29 per user per month. Research notes access to advanced models, including GPT-5, Claude, Gemini, and Mistral.

Does Dust support enterprise security controls?

Yes. Research notes AES-256 encryption at rest, audit logs, customer data ownership, and data residency options in the EU and US.

Does Dust offer audit logs?

Yes. The research states that audit logs are available, and workspace conversations can have custom data retention durations.

Who is Dust best for?

Research points to growth and scale-up teams of 20 to 500 people, especially non-technical users in sales, operations, and go-to-market roles. It is positioned less toward solo users or developers who want deep code-level control.

How does Dust compare with enterprise search tools?

Dust is positioned as more than enterprise search in the research, with a focus on agents that can automate workflows and take action on information. That differs from tools centered mainly on retrieval and search.

What are alternatives to Dust?

The research references Glean in Dust's own comparison content about enterprise search alternatives. More broadly, alternatives depend on whether a team wants search-first software or agent-building tools with stronger coding control.

Where can I watch dust?

Research does not point to a streaming service for dust.tt. Public references mention Dust videos and tutorials on YouTube, while the product itself is used through the web platform at dust.tt.