Comparison

VDF AI vs Dust

Dust is a team AI workspace that grounds assistants in Notion, Slack, Google Drive, and GitHub — built by former OpenAI and Stripe engineers, €29/user Pro, $40M Series B. VDF AI is the governed multi-agent orchestration platform for cross-system workflows with flat per-seat pricing and EU AI Act alignment. Here is an honest split by buyer intent.

Pick VDF AI if

You need agents that act across multiple enterprise systems (read and write), run multi-step DAG workflows, need EU AI Act evidence trails, or want flat per-seat pricing at governed production scale.

Pick Dust if

You want a fast, well-designed AI workspace where teams build assistants grounded in Notion, Slack, and Google Drive — and €29/user/month Pro or a custom Enterprise plan fits your budget and use case.

Book a Demo See VDF AI Agents
TL;DR

At a Glance

Four dimensions that drive most VDF AI vs Dust decisions.

Product shape
VDF AI
Governed multi-agent orchestration
Dust
Team AI knowledge workspace
Pricing
VDF AI
Flat per-seat enterprise
Dust
€29/user Pro · Enterprise custom (100+)
Agent capability
VDF AI
Multi-step DAG, write access, cross-system
Dust
Conversational assistants grounded in docs
Funding
VDF AI
Independent commercial platform
Dust
$40M Series B (2026) · Sequoia, Abstract
WHAT IS VDF AI?

An Enterprise AI Orchestration Platform

VDF AI targets platform teams accountable for governed production agents that go beyond answering questions — agents that take actions across enterprise systems, run multi-step workflows, and leave auditable evidence trails for compliance.

Networks v3 provides spec-driven DAG orchestration with nested networks. SEEMR (Self-Evolving Model Router) drives adaptive model selection. Vault persists encrypted runs. EU AI Act-aligned controls and data residency routing are in-product.

Agent Hub6-step builder, multi-provider routing, MCP tool registry, sandbox playground.
Networks v3Intent decomposition and nested networks for multi-agent production graphs.
SEEMRSelf-Evolving Model Router — four live dimensions and LinUCB modes. SEEMR architecture.
MCP ServerTool runtime with write access across enterprise SaaS — not just retrieval.
Vault + RBACCryptographically strong run history for investigations and compliance.
EU AI Act-alignedControls and residency paths for regulated industries.
WHAT IS DUST?

Team AI Knowledge Workspace

Dust was founded by former OpenAI and Stripe engineers and positions itself as a team AI workspace for building custom assistants grounded in company knowledge. Teams connect Notion, Slack, Google Drive, and GitHub as data sources; Dust assistants search across them using models from OpenAI, Anthropic, Google, and Mistral.

Pro plan is €29/user/month with a 14-day free trial. Enterprise requires a 100-user minimum and adds SSO (Okta, Entra, JumpCloud), SCIM provisioning, multi-workspace, US or EU hosting choice, and priority support. Dust holds SOC 2 Type II, GDPR, and HIPAA-enabling certifications. The company raised USD 40M Series B in May 2026 from Abstract and Sequoia with participation from Snowflake Ventures and Datadog.

Knowledge connectorsNotion, Slack, Google Drive, GitHub, and custom data sources as grounded context for assistants.
Custom AI assistantsBuild assistants with specific data sources, instructions, and model choices without writing framework code.
Multi-model supportOpenAI, Anthropic, Google, and Mistral — swap models per assistant without rewriting logic.
Team workspaceShared assistant library, role-based access, and collaborative assistant building across teams.
SSO + SCIM (Enterprise)Okta, Entra, JumpCloud, SCIM provisioning, multi-workspace, and priority support for 100+ user orgs.
EU hosting optionEnterprise plan includes choice of US or EU data hosting; SOC 2 Type II, GDPR, HIPAA-enabling.
SIDE BY SIDE

Feature by Feature

Dust pricing and features verified June 2026 against dust.tt/home/pricing and dust.tt product pages.

CapabilityVDF AIDust
Primary categoryGoverned multi-agent orchestration platformTeam AI knowledge workspace & assistant builder
PricingFlat per-seat enterprise€29/user/mo Pro; Enterprise custom (100+ users)
Knowledge base connectorsSemantic retrieval via MCP connectorsNotion, Slack, Google Drive, GitHub with native ingestion UX
Multi-step agentic workflowsNetworks v3 DAGs with write access, branches, nested agentsConversational assistants; complex multi-step DAGs: verify with Dust
Write access to enterprise systemsVia MCP tools (create, update, comment across SaaS)Retrieval-first; write actions: verify current Dust capabilities
Multi-model LLM supportAny provider via SEEMR adaptive routingOpenAI, Anthropic, Google, Mistral per-assistant choice
EU AI Act toolingIn-product aligned controls & data residency routingEU hosting option (Enterprise); GDPR certified; EU AI Act classification: not stated
Enterprise SSO / SCIMRBAC + enterprise authOkta, Entra, JumpCloud, SCIM on Enterprise plan
DeploymentCloud, hybrid, vendor-supported on-premCloud SaaS (US or EU hosting); no self-hosted edition
Cost & energy analyticsPer-node cost, latency, energy telemetryLLM cost visibility: verify current Dust capabilities
Target buyerEnterprise AI platform / risk teamsTeams wanting grounded AI assistants on company knowledge, fast

Dust pricing verified June 2026 against dust.tt/home/pricing. Dust is SaaS-only with US or EU hosting on the Enterprise plan; no self-hosted edition is available.

FAIR PLAY

Where Dust Wins

Dust earned strong traction with knowledge-work teams — here is where it shines.

Fast setup for knowledge assistants

Connect Notion, Slack, Google Drive, and GitHub in minutes; build assistants without writing framework code. The time from sign-up to a working, grounded assistant is measured in hours, not weeks.

Accessible per-user pricing

€29/user/month Pro makes Dust approachable for teams that don't yet need enterprise platform procurement — especially for startups and scale-ups with a self-service buying motion.

Collaborative assistant building

Team workspace with shared assistant library and role-based access means non-technical team members can build and improve assistants alongside developers — AI multiplayer in the daily workflow.

WHERE VDF AI WINS

When the Wedge Is Cross-System Agent Orchestration

VDF AI is built for agents that act, not just answer.

Multi-step agentic DAGs

Networks v3 lets agents branch, loop, and chain across multiple systems in a single orchestrated workflow — far beyond a single-turn knowledge retrieval assistant.

Write access to enterprise SaaS

MCP tools give agents create, update, and comment access across Microsoft 365, Jira, Confluence, GitHub, Slack, Zoom — not just search and retrieve.

EU AI Act alignment

Classification evidence, residency routing, and Vault audit trails are in the platform for European teams under Article 6–51 obligations — not just a hosting-region checkbox.

On-premises deployment

Vendor-supported on-prem deployment for regulated industries that cannot use SaaS-only platforms — Dust does not offer a self-hosted edition.

SEEMR adaptive routing

Adaptive multi-provider model routing optimising cost, quality, latency, and energy simultaneously — not just a per-assistant model dropdown.

DECISION GUIDE

Which One Should You Pick?

Separate “we need grounded answers from our knowledge base” from “we need agents that act across our software estate.”

Choose VDF AI if…

  • Your agents need write access to SaaS systems — creating Jira tickets, updating records, commenting in Slack.
  • You need multi-step DAG orchestration across multiple enterprise systems in one workflow.
  • EU AI Act evidence trails, data residency, or on-prem deployment are requirements.
  • Flat per-seat pricing at production scale is more predictable than per-user monthly billing.

Choose Dust if…

  • You want assistants that answer questions grounded in Notion, Slack, Google Drive, and GitHub — fast setup, no engineering overhead.
  • €29/user/month Pro or a custom Enterprise plan fits your buying motion and team size.
  • A collaborative team workspace where anyone can build and refine assistants is the primary UX requirement.
  • You don't yet need on-prem deployment or agent write access to production SaaS systems.

Using Dust for team knowledge today?

Keep Dust for the conversational knowledge layer your team already uses. When a workflow needs to write to Jira, trigger a GitHub action, loop through documents, or leave an EU AI Act audit trail, layer VDF AI for the orchestration step without replacing the assistants that are already working.

Plan the Agent Layer
FAQ

Frequently Asked Questions

What buyers ask when comparing VDF AI with Dust.

No. VDF AI is an independently built enterprise AI orchestration platform with Agent Hub, Networks v3, MCP Server, Vault, and SEEMR — a Self-Evolving Model Router for adaptive governed routing across any LLM provider. Dust is an AI workspace platform founded by former OpenAI and Stripe engineers that lets teams build custom AI assistants grounded in company knowledge bases (Notion, Slack, Google Drive, GitHub). Dust raised USD 40M Series B in 2026 (Abstract, Sequoia, Snowflake Ventures, Datadog). The two are independent commercial products with different scopes.

Dust (verified June 2026 on dust.tt/home/pricing): Pro is €29 per user per month after a 14-day free trial, for small teams and startups. Enterprise requires custom pricing and a minimum of 100 users; it adds SSO (Okta/Entra/JumpCloud), SCIM, multi-workspace, US/EU hosting choice, and priority support. Dust holds SOC 2 Type II, GDPR, and HIPAA-enabling certifications. VDF AI is flat per-seat commercial pricing that bundles runtime, multi-agent orchestration, enterprise connectors, observability, and EU AI Act governance in one platform fee.

A Dust assistant is a knowledge-grounded conversational agent scoped to a team's connected data sources — Notion pages, Slack messages, Google Drive files, GitHub repos. It answers questions and surfaces insights from those sources using models from OpenAI, Anthropic, Google, or Mistral. A VDF AI agent runs inside a Networks v3 DAG: it can read from and write to enterprise SaaS systems, call external APIs via MCP tools, route across models with SEEMR, and chain into multi-agent workflows — going beyond retrieval to action and cross-system orchestration.

Yes. Teams sometimes use Dust for the everyday conversational assistant layer — answering questions grounded in Notion or Slack — and VDF AI for complex multi-step workflows that span multiple systems, require write access, or need EU AI Act evidence trails. VDF AI can call Dust agent endpoints via its MCP tool layer if Dust exposes an API, or the two run independently at different layers of the AI toolchain.

Dust Enterprise offers a choice of US or EU hosting, with SOC 2 Type II and GDPR certifications. VDF AI provides EU AI Act-aligned controls, data residency routing at the agent and workflow level, and Vault audit evidence that goes beyond hosting choice — purpose-built for European regulated enterprises that must demonstrate Article 6–51 compliance at the orchestration layer, not just at the infrastructure level.

When your primary need is a team knowledge assistant — connecting Notion, Slack, Google Drive, and GitHub so employees can ask questions and get grounded answers — and €29/user/month Pro pricing fits your budget. Dust excels at rapid setup, strong knowledge-source UX, and multi-model flexibility for teams that don't need cross-system write access or multi-agent DAG orchestration. VDF AI is the stronger fit when agents must take actions across systems, workflows span multiple enterprise applications, or EU AI Act evidence and flat per-seat enterprise pricing are primary requirements.
EXPLORE MORE

Related Resources

Related Comparisons
Related Resources
VDF AI Products

Ready for agents that act, not just answer?

When your knowledge assistants need to write to Jira, update records, or run EU AI Act-auditable multi-step workflows, VDF AI picks up where knowledge workspaces leave off.

View feature comparison