# Dust — AI Agents & Assistants

Dust by Permutation Labs holds a panel rating of 3.5/5 from 6 AI judges (category median 3/5), as of 2026-10-28. Provenance: EU-Made; sovereignty: 1 of 4 dimensions proven.

- Vendor: Permutation Labs (https://dust.tt)
- Category: [AI Agents & Assistants](https://whats-best.ai/software/ai-agents.md)
- Provenance (editorial classification): EU-Made
- Canonical page: https://whats-best.ai/software/ai-agents/dust
- Alternatives: https://whats-best.ai/software/ai-agents/dust/alternatives ([markdown](https://whats-best.ai/software/ai-agents/dust/alternatives.md))
- Data as of: 2026-10-29 · evaluated 2026-10-28 · oldest source captured 2026-10-26

**AI-generated.** Every score, rationale and verdict here is written by a panel of AI judges from public evidence — an analysis, not a hands-on test. Every fact quotes the vendor page it was read from. No score can be bought. How scores are made: https://whats-best.ai/methodology

## At a glance

- Vendor: Permutation Labs
- Headquarters: Not stated on the vendor’s pages as of 26 Oct 2026
- Data hosting: Default not stated; pages mention the US and the EU [S1]
- Free plan: Yes, per the vendor’s own page [S2]

## The panel's verdict

Dust, an agent platform from Permutation Labs of Paris, is strongest on integrations and knowledge, where the computed range is 8 to 8: 20+ connectors from Slack and Notion to Salesforce and Zendesk, MCP in both directions, and grounded answers that link back to their sources. Processing and models follows at 7-8 — providers are named, and an EU region at eu.dust.tt can lock the model picker, agent builder and auto models to EU-hosted models, refusing activation while any agent still points at a non-EU model. The lowest ranges are agent capabilities and sovereignty, both 6-7, where lower scores cite human approval described only in general terms, no public information on test-and-evaluation tooling, acknowledged transfers to the United States under standard contractual clauses, and no public information on subprocessor locations. The flagged splits came back empty, and no criterion spans more than a single point; persona-weighted totals run 6.8 to 7.6. Pricing transparency holds at 7: without published per-model credit rates, overage figures or enterprise pricing, annual cost is not computable from public pages.

**Speaks for it:**
- Managed connectors cover Slack, Notion, Confluence, Drive, Gmail, Outlook, HubSpot, Salesforce, Zendesk and more, with data scoping down to folders, sub-pages and specific channels.
- An EU-hosted-models setting restricts the model picker, agent builder and auto models, and refuses to activate while any agent still uses a non-EU model.
- Written commitments that data never trains models, extended by name to every model provider under zero data retention, with SOC 2 Type II published.
- SSO, automated provisioning and deprovisioning, SCIM-synced groups, three roles with additive permissions and audit logs are all documented.
- Public seat prices carry credit allocations — Pro 30€/month or 24€/month billed yearly with 8,000 credits/month, Max 150€/month or 120€/month billed yearly with 40,000 credits/month.

**Held against it:**
- We found no public information on exporting data, agents and configuration when leaving the platform.
- We found no public information on exporting audit logs or streaming them to a SIEM.
- We found no public information on per-action human approval, on test-and-evaluation tooling before release, or on the handling of agent failures.
- We found no public information on per-model credit consumption, the overage cap and rate, or enterprise volume pricing, so annual cost is not computable from public pages.
- The privacy policy acknowledges transfers of personal data to the United States under standard contractual clauses, we found no public information on subprocessor locations, and the EU-hosted-models setting is documented as incompatible with bring-your-own-key.

**Best for:**
- You need agents that search, create and update across company tools like Notion, Confluence, GitHub, Salesforce, HubSpot and Zendesk, with answers that link back to their sources.
- Your workspace must run in the EU with model choice pinned to EU-hosted models and admin caps on which models members may select.
- Your legal review starts from written no-training commitments, zero data retention for named providers, SOC 2 Type II and an enterprise DPA.
- Your IT team needs documented SSO, SCIM provisioning, additive role permissions and audit logs before rollout.

**Avoid if:**
- You require a fully European processing chain end to end — transfers of personal data to the United States under standard contractual clauses are acknowledged, and we found no public information on subprocessor locations or company ownership.
- You want EU-hosted models while supplying your own model provider keys — the two are documented as incompatible.
- You need a computable annual or per-conversation cost for procurement — ask the vendor: the public pages we read do not show it
- You need per-action human approval before agents act in connected systems — the captured pages describe human review only in general terms.

## Scores by judge and criterion

Scores out of 5, each judge scoring independently against the published rubric (v1). "n/r" = not rated: the vendor publishes no price, or the judge found no public information on that criterion.

| Criterion | The Head of IT | The Data Protection Officer | The Automation Lead | The Customer-Service Director | The Works Council Member | The Skeptic |
| --- | --- | --- | --- | --- | --- | --- |
| Data use, training & retention | 3.5/5 | 3.5/5 | 4/5 | 4/5 | 3.5/5 | 3.5/5 |
| Models & processing location | 4/5 | 4/5 | 4/5 | 4/5 | 4/5 | 3.5/5 |
| Admin, identity & access controls | 3.5/5 | 3.5/5 | 4/5 | 3.5/5 | 4/5 | 3.5/5 |
| Integrations, connectors & knowledge | 4/5 | 4/5 | 4/5 | 4/5 | 4/5 | 4/5 |
| Agent capabilities & guardrails | 3.5/5 | 3.5/5 | 3.5/5 | 3/5 | 3.5/5 | 3/5 |
| European sovereignty | 3.5/5 | 3.5/5 | 3.5/5 | 3.5/5 | 3.5/5 | 3/5 |
| Pricing transparency | 3.5/5 | 3.5/5 | 3.5/5 | 3.5/5 | 3.5/5 | 3.5/5 |
| **Weighted verdict** | **3.5/5** | **3.5/5** | **4/5** | **3.5/5** | **3.5/5** | **3.5/5** |

### The judges

- **The Head of IT** (v1.0): Has to roll an AI assistant out to eight hundred staff without a second shadow-IT wave. Wants SSO and SCIM, groups that decide who gets which model, agent and connector, answers that respect SharePoint permissions, audit logs for the SIEM, and a deployment option that is not "trust us".
- **The Data Protection Officer** (v1.0): Signs the DPA and gets asked where a prompt goes. Wants every model provider named as a subprocessor with its region, training excluded for all of them, retention the company controls, and an EU-only switch that also covers agents and connectors — not just the chat window.
- **The Automation Lead** (v1.0): Builds the agents that file tickets, draft offers and chase approvals across the ERP, CRM and wiki. Wants tools and actions with scoped permissions, MCP and an API, approval steps before anything is sent, run logs, and a way to test an agent before the whole company uses it.
- **The Customer-Service Director** (v1.0): Runs a contact centre with forty agents and wants half the routine calls and chats answered by an AI agent. Wants handover to a human with the full context, actions in the order and ticket systems, evaluation before go-live, and a price per conversation the CFO can forecast.
- **The Works Council Member** (v1.0): Has a co-determination right the moment a tool can monitor staff, and an AI assistant logs every prompt. Wants usage reports that can be aggregated or pseudonymised, prompt logging rules set out in writing, a human approving what agents do to colleagues, and no training on staff input.
- **The Skeptic** (v1.0): Has read "European AI" on a wrapper that sends every prompt to a US model, "agents" that are saved prompts, and "we never train on your data" with no word on the model provider behind it. Asks which model answers, where it runs, what it may do unsupervised, and what a thousand seats cost.

## European sovereignty — proven facts

Built only from facts on the vendor's own pages; a dimension we could not prove is left open, never scored as zero. Headline: 1 of 4 dimensions proven.

| Dimension | Finding | Points | Source |
|---|---|---|---|
| Legal entity | Not determined | — | — |
| Ownership | Not determined | — | — |
| Data residency | Not determined | — | — |
| Subprocessors | US CLOUD Act reach | 0/2 | [S4] |

## Key facts

Each fact is the vendor's own statement, cited to the source page it was read from ([S*n*] → Sources).

**Compliance**
- Audit logs: yes · yes [S1]
- California rights: yes · 1798.83 [S4]
- Childrens privacy: 16 [S4]
- Cookies posthog: yes · yes · yes · yes [S2]
- Encryption: AES-256 · TLS [S3]
- GDPR SOC 2 HIPAA: yes · yes · yes [S3]
- International transfer: adequacy decision · Standard Contractual Clauses · United States · yes [S4]
- Permissions model: yes · yes · yes · yes · yes [S1]
- Privacy contact: privacy@dust.tt [S4]
- Provisioning: yes [S1]
- Security measures: yes · yes [S4]
- SOC 2: yes [S1]
- SOC 2: yes [S2]
- SSO: yes [S1]

**Data**
- Retention: yes [S1]
- Retention contractual: platform delivery, contract-related requests · agreement duration + 5 years archive [S4]
- Retention events: event management · event duration + 5 years archive [S4]
- Retention marketing: Marketing Communications · 3 [S4]

**Hosting**
- Data residency: US · EU [S1]
- Data residency: US · EU · yes [S2]
- EU activation check: yes [S5]
- EU availability requirements: no · yes · yes [S5]
- EU models setting: EU-hosted models only setting restricts agents to EU-hosted models [S5]
- EU region URL: https://eu.dust.tt · yes [S5]
- EU setting path: Admin > Models [S5]
- Providers: PostgreSQL · Google Cloud Platform · QDrant Cloud [S4]
- Regions: EU · US [S3]
- Single tenant: yes [S1]
- Single tenant: yes [S2]

**Integrations**
- Automation platforms: Zapier · Make · n8n · Power Automate [S2]
- Developer API: developer API · conversation API + data source API [S2]
- Google drive: yes · Google Docs · Google Slides · pdfs · plain text files · yes · yes [S4]
- Workspace management mcp tools: list_groups · get_group_members · create_group · update_group_members [S8]

**Legal**
- Custom terms: yes · yes · yes [S2]
- Entity name: 86 avenue de Wagram, 75017 Paris, France · Permutation Labs · France · Dust [S4]
- Policy update notice: yes [S4]

**Pricing**
- Annual discount: 20 [S2]
- Business team size: 100 [S2]
- Business yearly max: 40000 · 120 [S2]
- Business yearly pro: 8000 · 24 [S2]
- Credit rollover: no [S2]
- Enterprise features: Enterprise deployment includes shared workspace credits and volume pricing [S1]
- Free tier: 0 · 500 [S2]
- Max monthly: 150 · 40000 · 120 [S2]
- Midcycle proration: yes [S2]
- Overage: yes · yes · yes [S2]
- Pro monthly: 30 · 8000 · 24 [S2]
- Seat reassignment: yes · yes [S2]
- Trigger credits: yes · member · workspace [S12]
- VAT: yes [S2]

**Product**
- Access determination: workspace role · group permissions · resource access · yes [S8]
- Admin model caps: models · reasoning efforts · auto models per member · yes [S6]
- Admin role capabilities: yes · yes · yes · yes [S8]
- Agent access policy: users create agents only from data of spaces they have access to [S7]
- Agent distribution: agents can be deployed company-wide and distributed across groups [S7]
- Agent memory tool: Agent Memory · yes [S11]
- Audit logs toggle: yes [S8]
- Auto model updates: periodic benchmarking and updates; first available candidate used per message [S6]
- Auto models: ordered list of model and reasoning-effort combinations maintained by Dust · Basic · Standard · Premium [S6]
- Automations workspace credits: triggers run on workspace credit pool instead of member credits [S8]
- Chrome extension: yes [S1]
- Collaboration workspaces: yes [S2]
- Company knowledge context: Agents use context across Slack, CRM, docs, and other company systems [S1]
- Computer tool: Computer · Word · Excel · PowerPoint · PDF · yes [S11]
- Confluence tool: Confluence · search, create, and update pages · Personal or Workspace [S11]
- Connections: Slack · Notion · Google Drive [S7]
- Connections scope: folder · document [S7]
- Connectors: Slack · Notion · Google Drive · GitHub · Salesforce · Zendesk · 70+ [S1]
- Connectors: yes · 20+ [S2]
- Credit definition: credit is unit for measuring AI usage, consumed per message based on model and actions [S2]
- Custom agents: yes [S2]
- Customer count: 3000+ [S1]
- Data sources: yes · yes · yes · job position · job title · workplace · industry · name · email · telephone [S4]
- Data visualization tool: Data visualization · yes [S11]
- Datasource prereq: Connections, Folders and/or websites must exist as data sources [S10]
- Default tools: yes · no [S11]
- Dust agent all sources: @dust agent can query all data sources if workspace is configured with it [S10]
- Dust apps deprecated: Dust Apps let agents execute custom code as tools, deprecated [S11]
- Enterprise pooled credits: yes [S2]
- EU flag indicator: EU flag in model picker [S5]
- EU hosted models anthropic: Claude Opus 5.5 · Claude Opus 5 · Claude Opus 4.8 · Claude Sonnet 5.5 · Claude Sonnet 5 · Claude Sonnet 4.6 · Claude 4.5 Haiku · Anthropic [S5]
- EU hosted models fireworks: GLM-5.3 · Fireworks [S5]
- EU hosted models google: Gemini 3.8 Flash · Gemini 3.7 Flash · Google [S5]
- EU hosted models mistral: Mistral Large 4 (Preview) · Mistral Large · Mistral Medium 3.5 · Mistral Small · Mistral Codestral · Mistral [S5]
- EU hosted models openai: GPT 6.1 Sol · GPT 6 Sol · GPT 6 Astra · GPT 6 Luna · GPT 5.6 Sol · GPT 5.6 Terra · GPT 5.6 Luna · GPT-5.4 Mini · GPT-5.4 Nano · OpenAI [S5]
- EU model restriction effect: Basic · Standard · Premium · yes [S5]
- Filesystem navigation: yes [S10]
- Flexible model providers: Flexible Providers [S3]
- Frame sharing policy: yes [S8]
- Frames: Frames: interactive dashboards & apps, Standard on Business, White-labelled on Enterprise [S2]
- Github tool: GitHub · search, update, comment, create pull requests and issues · Workspace [S11]
- Gmail tool: Gmail · search emails and create drafts · Personal [S11]
- Google calendar tool: Google Calendar · search, update and create events · Personal [S11]
- Governance: activity · adoption · usage · costs · ROI · yes · yes [S1]
- Governance admin only settings: pod and feature settings · governance permissions [S8]
- Granular data selection: Granular Data Selection [S3]
- Granularity recommendation: constrain agent data sources for better performance [S10]
- Group managers: membership, credit limits, credit requests · yes [S8]
- Group types: provisioned (SCIM) · manual [S8]
- Hubspot tool: HubSpot · search, update and create info · Personal [S11]
- Image generation tool: Create Images · yes [S11]
- In conversation filter exact match: yes [S10]
- Label filter modes: exclude specific content · include specific content · in-conversation dynamic filtering [S10]
- Label filter reset warning: removing and re-adding a knowledge configuration loses existing must-have/must-not-have label filters [S10]
- Label filtering: yes [S10]
- Label support scope: Folders and most Connections [S10]
- Matrix access billing: yes_by_default · when_granted · when_granted [S8]
- Matrix create agents: yes · when_granted · when_granted [S8]
- Matrix invite by email: yes_when_policy_allows · when_granted_if_policy_allows · when_granted_if_policy_allows [S8]
- Matrix model providers: yes · no · no [S8]
- Matrix share public link: yes_when_policy_allows · when_granted_if_policy_allows · when_granted_if_policy_allows [S8]
- Matrix view analytics: yes · no · yes [S8]
- Member role access: no [S8]
- Model access tiers: Some models require an included model access tier [S5]
- Model flexibility: OpenAI · Anthropic · Google · Mistral · 20+ [S1]
- Model selection locations: model picker in composer · Advanced > Model selection in Agent Builder [S6]
- Models: 20+ frontier models: GPT, Claude, Gemini, Mistral, DeepSeek [S2]
- Multi agent workflows: yes [S2]
- Multiplayer: Multiplayer AI: shared context, reusable skills, connected tools, human review [S1]
- New workspace defaults: Create agents · Publish agents · Invite people by email · Share by public link · Create skills · Manage skill availability · Make skills discoverable to agents · Charge automations to the workspace [S8]
- No capability behavior: agent with no capability uses only the model [S11]
- No model training: No Model Training [S3]
- Not EU hosted models: Grok 4.7 · Grok 4.6 · Gemini 3.1 Pro (Preview) · Gemini 3.5 Flash Lite · Gemini 3.1 Flash Lite · DeepSeek V4.1 Flash · Kimi K3 · GLM-5.3 Flash · Inkling [S5]
- Notion tool: Notion · search, update and create pages · Workspace [S11]
- Outlook tool: Outlook · search emails and create drafts · Personal [S11]
- Permission modes: Everyone · Groups · Admin only [S8]
- Permissions additive: yes [S8]
- Pods: yes [S2]
- Positioning: Multiplayer AI for human-agent collaboration; agents over company knowledge, tools and teams [S1]
- Private spaces: Private Spaces [S3]
- Providers: OpenAI · Anthropic · Google · Mistral [S6]
- Rag search: agents can semantically search selected data sources via RAG before sending the message to the model [S10]
- Remote mcp servers: remote MCP servers can be added as tools [S11]
- Retrieval response links: yes [S10]
- Role based access: Role-Based Access [S3]
- Roles: user roles: Members, Builders, Admins [S7]
- Run agent tool: Run an agent · yes [S11]
- Salesforce tool: Salesforce · search, update and create info · Personal [S11]
- Sharepoint default nav: yes [S10]
- Skill discoverability permission: yes [S8]
- Slack tool: Slack · search info and post messages · Personal [S11]
- Space membership: admins define space membership [S7]
- Spaces: open · restricted [S7]
- SSO: Single Sign-On (SSO) [S3]
- SSO: Okta · Entra ID · Jumpcloud · 5 [S2]
- Subpage selection: connections · folders · websites · sub-pages · specific channels [S10]
- Supported sync sources: /Google Drive · /Confluence · /Notion · /Slack [S10]
- Third party tools admin config: Personal · Workspace · yes [S11]
- Tools default access: workspace-wide [S11]

**Subprocessors**
- Foundation model providers: , OpenAI · Anthropic · Mistral · Google · Fireworks · yes [S4]
- Zero data retention: yes [S4]

**Support**
- Enterprise: yes · yes [S1]
- Enterprise tier: email · priority + SLA, dedicated CSM & onboarding [S2]
- Trust center: Trust datasheet and Trust Center available [S1]
- Trust center: Trust Center available [S3]

19 more facts on the canonical page: https://whats-best.ai/software/ai-agents/dust

## Questions

### Is Dust European?

Yes. whats-best.ai classifies Dust as EU-Made.

### Where does Dust store data?

The default storage location for customer data is not stated on the vendor’s pages as of 26 Oct 2026. The vendor’s pages mention the US and the EU as hosting locations, per [the vendor’s page captured 26 Oct 2026](https://whats-best.ai/evidence/10787).

### Does Dust have a free plan?

Yes. Dust has a free plan, per [the vendor’s page captured 26 Oct 2026](https://whats-best.ai/evidence/10788).

### What are the best alternatives to Dust?

As of 28 Oct 2026, the panel rates [Langdock (4/5)](https://whats-best.ai/software/ai-agents/langdock), [Unique AI (3.5/5)](https://whats-best.ai/software/ai-agents/unique-ai) and [Omnifact (3.5/5)](https://whats-best.ai/software/ai-agents/omnifact) highest among the other 15 AI Agents & Assistants products evaluated. Ratings that look equal are rounded to half stars; the order uses the unrounded rating, then the name. The highest-rated European alternative is [Langdock](https://whats-best.ai/software/ai-agents/langdock) (EU-Made). Ratings are the AI panel’s opinion, formed from public evidence: [how scores are made](https://whats-best.ai/methodology#ordering). The full list, with European products first: [Dust alternatives](https://whats-best.ai/software/ai-agents/dust/alternatives).

## What this page leaves out

- 11 integrations facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material.
- 5 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material.
- 4 compliance facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material.
- 1 legal fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 1 pricing fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 2 sovereignty dimensions could not be confirmed on the vendor’s own pages and are shown as unknown.

## Sources

- [S1] Vendor homepage: https://dust.tt/ (checked 2026-10-26)
- [S2] Pricing page: https://dust.tt/home/pricing (checked 2026-10-26)
- [S3] Security page: https://dust.tt/home/security (checked 2026-10-26)
- [S4] Privacy policy naming the legal entity: https://dust.tt/home/platform-privacy (checked 2026-10-26)
- [S5] Models & processing location — found from sitemap: https://docs.dust.tt/docs/user-documentation/admins/admin-governance/eu-hosted-models (checked 2026-10-26)
- [S6] Models & processing location — found from sitemap: https://docs.dust.tt/docs/user-documentation/getting-started/faq/general-questions/what-ai-models-are-available-in-dust (checked 2026-10-26)
- [S7] Admin, identity & access controls — found from sitemap: https://docs.dust.tt/docs/user-documentation/admins/admin-governance/access-controls-and-permissions (checked 2026-10-26)
- [S8] Admin, identity & access controls — found from sitemap: https://docs.dust.tt/docs/user-documentation/admins/admin-governance/workspace-governance-roles-groups-and-permissions (checked 2026-10-26)
- [S9] Integrations, connectors & knowledge — found from sitemap: https://docs.dust.tt/docs/user-documentation/agents/integrations/dust-mcp-server (checked 2026-10-26)
- [S10] Integrations, connectors & knowledge — found from sitemap: https://docs.dust.tt/docs/user-documentation/agents/knowledge/search-data-sources (checked 2026-10-26)
- [S11] Agent capabilities & guardrails — found from sitemap: https://docs.dust.tt/docs/user-documentation/agents/tools (checked 2026-10-26)
- [S12] Agent capabilities & guardrails — found from sitemap: https://docs.dust.tt/docs/user-documentation/agents/triggers (checked 2026-10-26)

---

Cite as: whats-best.ai, "Dust — AI Agents & Assistants", AI panel analysis, as of 2026-10-29. https://whats-best.ai/software/ai-agents/dust

Corrections: https://whats-best.ai/correct/1029?target=page