Dust and Gemini Enterprise are both platforms where a team builds and runs AI agents on company data, but they start from different places. Dust is model-neutral: its Business plan offers “20+ frontier models” from OpenAI, Anthropic, Google, Mistral and DeepSeek, with seats from 0€ to 120€ per seat per month billed yearly, excluding VAT, and US or EU data residency. Gemini Enterprise is Google’s agent platform: Business edition starts at $21 per seat per month with a 30-day trial, and Standard and Plus add stricter compliance, data residency and bring-your-own agents. On memory, both document personal memory per user. Dust also documents shared Pods that agents use as context, and skills that propose improvements from user corrections, applied after an editor approves. Google describes memory for tasks, facts and methods; the Gemini pages we read do not show how one colleague’s correction reaches other colleagues’ agents.
By Bombos. Published November 17, 2026. Sources checked November 17, 2026.
What are Dust and Gemini Enterprise?
Dust calls itself “multiplayer AI”. A workspace holds agents, company data and people. Admins add data through Connections, such as Google Drive, Notion, Slack, Confluence, GitHub, Microsoft, Zendesk and BigQuery, which “sync external platforms into Dust so that agents can search and retrieve your company’s knowledge” (Dust docs, Connections). Data lives in Spaces: Company Data is “the default Space of your workspace”, and “only members of the Spaces the agent uses can see and use it” (Dust docs, Quickstart).
Gemini Enterprise is, in Google’s words, “an advanced agentic platform that brings the best of Google AI to every employee”. Employees use prebuilt agents such as Deep Research and Gemini Notebook (NotebookLM) and build their own with “a powerful no-code workbench” (Gemini Enterprise FAQ). It lists “over 100 supported integrations” (Gemini Enterprise overview). Connectors work in two ways: “Data ingestion (indexing) copies data into the Gemini Enterprise index”, while “Data federation directly retrieves information from the specified data source” (Google Cloud, Connectors and data stores).
If you are still deciding whether you need Gemini Enterprise at all, or Gemini inside Google Workspace is enough, read Gemini Enterprise vs Gemini in Google Workspace first.
How do Dust and Gemini Enterprise compare?
| Aspect | Dust | Gemini Enterprise |
|---|---|---|
| Models | “20+ frontier models: GPT, Claude, Gemini, Mistral, DeepSeek”, chosen per agent | “Google Gemini models today, with additional models coming soon”; third-party model choice is in early access |
| Building agents | Agent builder with instructions, tools and knowledge; “multi-agent workflows on schedules & triggers” | No-code Workflow Builder in Business; Standard and Plus can bring agents built on Agent Platform or ADK, or connect external agents through A2A |
| Company data | Connections synced into Spaces; “Connect Slack, Notion, GitHub, Drive + 20 more, or any tool via MCP” | Connectors such as Google Workspace, Microsoft 365, ServiceNow, Jira and Confluence; data is either ingested into the Gemini Enterprise index or retrieved through federated search |
| Access control | Agents and data follow Space membership | Runs under each person’s existing permissions, with document-level access controls (Standard and Plus) |
| Memory | Agent Memory per user and per agent; shared Pods whose conversations agents use as context; self-improving skills that propose changes from feedback and corrections, applied after editor approval | Personal Memory and personalization, on by default, admins can turn it off; overview names session, semantic, procedural and episodic memory |
| Data residency | “US & EU data residency” on Business; single-tenant deployment on Enterprise | us, eu and global locations; the FAQ lists data residency under Standard and Plus |
| Certifications | “SOC2 Type II Certified”, “GDPR Compliant” | SOC 1, 2 and 3 listed for Standard, Plus and Gemini Notebook Enterprise |
| Published price | Free 0€, Pro 24€, Max 120€ per seat per month billed yearly, excluding VAT | Business “Starting at $21 per seat per month”; Standard and Plus through Google Cloud sales |
| Trial | Free seat with “500 credits, Lifetime” | 30 days free for Business, Standard or Plus |
What does each one cost?
Dust. The Business plan is “for teams up to 100 people” and is priced per seat. Free is “0€, 500 credits lifetime”. Pro is “30€/month, or 24€/month billed yearly, with 8,000 credits/month”. Max is “150€/month, or 120€/month billed yearly, with 40,000 credits/month”. Prices exclude VAT. Unused credits do not roll over. Enterprise is priced through sales and adds unlimited connectors and MCP servers, SCIM, audit logs and single-tenant deployment (Dust pricing).
Gemini Enterprise. The Business edition page says “Starting at $21 per seat per month” (Gemini Enterprise Business). The FAQ says the app “is billed on a per-seat, per-month basis” and refers to Google Cloud sales for buying Standard, Plus or Frontline. We found no published per-seat price for Standard or Plus on the Google pages we read. The FAQ describes Business as for “teams up to 300”, while the editions comparison lists Business as “1-500 users” (Compare editions).
Usage counts in both: Dust in credits, Gemini Enterprise in quotas and pay-as-you-go usage.
How do memory and shared knowledge work in each?
Dust has an Agent Memory tool. “The memory tool enables agents to accumulate and retain information about users across conversations.” It is scoped per user and per agent: “Memories are private to each user and only you can view and manage your memories with each agent” (Dust docs, Agent Memory). Shared knowledge in Dust comes from three places. First, the data you connect to Spaces. Second, Pods: a Pod is a “shared workspace inside Dust where a team (humans and agents together) collaborates around a common goal”, with conversations, tasks and files, and its conversations are “automatically indexed so agents can reference them as context” (Dust docs, Pods). Third, self-improving skills: Dust reviews conversations nightly and uses signals such as “a user correcting the agent’s output” to propose edits to a skill. “Nothing is applied to your skill without your explicit approval” (Dust docs, Self-improving skills).
Gemini Enterprise builds a “personal Memory by understanding your individual needs and work patterns”. Memory is on by default, and administrators “can disable Memory and personalization profiles for all users” (Google Cloud, Personalization). Google’s overview names four kinds: “Session memory for active tasks, semantic memory for structured business knowledge, procedural memory for workflows, and episodic memory of past actions across devices and channels” (Gemini Enterprise overview). The personalization page describes Memory as personal and does not say where memory data is stored. The pages we read do not show how a correction made by one colleague becomes shared knowledge for other colleagues’ agents.
In short: both platforms remember things per person and give the team shared access to connected documents. Dust adds shared Pods and skills that improve from corrections after an editor approves them. For Gemini Enterprise, ask Google how a correction such as “this price list replaced the old one” reaches every colleague’s agent. Why this matters is explained in Memory layer for AI agents.
Where is the data, and which certifications apply?
Dust says on its security page “Host in the EU or US to meet your regulatory needs”, “SOC2 Type II Certified”, “GDPR Compliant” and “Your data is never used to train models” (Dust security). EU hosting means “data is stored on servers located within the European Union”, with “model inference routed to EU regions where available” (Dust blog, data sovereignty). Not every model runs in the EU: the changelog says Grok 4.7 “is available to everyone except workspaces configured for EU data residency only” (September 22, 2026) (Dust changelog).
Gemini Enterprise offers us, eu and global locations. In the eu multi-region you can keep customer data at rest in the EU, and machine learning processing stays in the EU “when requests are made to regionally-available APIs”. Traffic on global “doesn’t support regional data residency”, and grounding with Google Search is only available in global (Google Cloud, Locations). The FAQ lists data residency, audit logging, customer-managed encryption keys and VPC controls under Standard and Plus (Gemini Enterprise FAQ). Google’s compliance page lists SOC 1, SOC 2 and SOC 3 for Standard, Plus and Gemini Notebook Enterprise (Google Cloud, Compliance). We did not find a statement on SOC 2 for the Business edition.
Which one fits a 10 to 200 person team?
- Your team uses models from several vendors. Dust lets you pick the model per agent. Gemini Enterprise runs Gemini models today.
- Your work lives in Google Workspace. Gemini Enterprise connects Drive, Gmail, Calendar and Groups “with real-time syncing” and includes Gemini Notebook.
- You need EU data residency at a small team size. Dust lists US and EU residency on its self-serve Business plan. For Gemini Enterprise, the FAQ places data residency under Standard and Plus, which are bought through sales.
For a broader look at Gemini Enterprise next to Bombos, see Google Gemini Enterprise or Bombos.
Where does Bombos fit?
Bombos is not a third agent platform. It is the memory layer under the AI a team already uses: one shared memory that learns how the team works, with every answer showing what it rests on.
- It sits under your AI, not next to it. Bombos is a plugin and does not act itself: your AI acts, Bombos remembers. Bombos works with Claude, ChatGPT, Copilot, Grok, Codex, Cursor and your own agents. Dust and Gemini Enterprise are not on that list today: compatibility with either has not been demonstrated.
- One shared memory for the team. Bombos keeps what the team settled, for everyone, with its source, across the AI tools it supports.
- It runs in the cloud on servers in Europe, or standalone on your own server.
- One person can start. Try the plugin first for your own agents; the team joins when it works for you.
Frequently asked questions
Is Dust cheaper than Gemini Enterprise? It depends on seat mix and usage. Dust Pro is 24€ per seat per month billed yearly, excluding VAT, with 8,000 credits. Gemini Enterprise Business starts at $21 per seat per month. Standard and Plus prices come from Google Cloud sales.
Does either platform have team memory? Both document personal memory. Dust also documents shared Pods and self-improving skills. The Gemini pages we read do not show how a correction becomes shared across colleagues’ agents.
Can I keep data in the EU?
Dust offers EU hosting on Business. Gemini Enterprise offers the eu multi-region; the FAQ lists data residency under Standard and Plus.
Book your half hour: together we map what your AI forgets. Free, and the plan is yours, with or without Bombos.
Which sources did we use?
All sources opened and read on November 17, 2026.
- Dust, Pricing: https://dust.tt/home/pricing
- Dust, Security: https://dust.tt/home/security
- Dust docs, Agent Memory: https://docs.dust.tt/docs/agent-memory
- Dust docs, Quickstart: https://docs.dust.tt/docs/user-documentation/admins/quickstart
- Dust docs, Connections: https://docs.dust.tt/docs/user-documentation/admins/connections-management
- Dust docs, Pods: https://docs.dust.tt/docs/user-documentation/pods/overview
- Dust docs, Self-improving skills: https://docs.dust.tt/docs/user-documentation/agents/self-improving-skills
- Dust docs, Tools: https://docs.dust.tt/docs/tools
- Dust docs, Changelog: https://docs.dust.tt/docs/changelog
- Dust blog, What is data sovereignty: https://dust.tt/blog/what-is-data-sovereignty
- Google Cloud, Gemini Enterprise overview: https://cloud.google.com/gemini-enterprise
- Google Cloud, Gemini Enterprise Business edition: https://cloud.google.com/gemini-enterprise/business
- Google Cloud, Gemini Enterprise app FAQs: https://cloud.google.com/gemini-enterprise/faq
- Google Cloud docs, Compare editions: https://docs.cloud.google.com/gemini/enterprise/docs/editions
- Google Cloud docs, Personalization and Memory: https://docs.cloud.google.com/gemini/enterprise/docs/configure-personalization
- Google Cloud docs, Locations: https://docs.cloud.google.com/gemini/enterprise/docs/locations
- Google Cloud docs, Compliance and security controls: https://docs.cloud.google.com/gemini/enterprise/docs/compliance-security-controls
- Google Cloud docs, Introduction to connectors and data stores: https://docs.cloud.google.com/gemini/enterprise/docs/connectors/introduction-to-connectors-and-data-stores
