Key takeaways
- Real revenue at scale: Glean surpassed $200M ARR in December 2025 — doubling in nine months — and raised a $150M Series F at a $7.2B valuation in June 2025
- Glean Agents is a no-code builder plus shared Agent Library deployed over a permissions-aware knowledge graph spanning 27B+ indexed documents and 100+ connectors, with 250M+ agent actions executed
- Multi-model by design: the Model Hub lets builders pick OpenAI, Anthropic, Google, Amazon, or Meta models per agent and per workflow step — but the platform is closed source, enterprise-sales only, with no public pricing
FAQ
What is Glean?
Glean is an enterprise Work AI platform that indexes a company's knowledge across 100+ applications with permissions awareness, then layers an AI assistant and Glean Agents — shared, no-code AI agents — on top, accessible from Slack, Microsoft Teams, the web, and an API.
How much does Glean cost?
Pricing is not publicly listed. Glean sells through enterprise sales with custom quotes; third-party analyses estimate roughly $35–50 per user per month with minimum commitments of around 100 seats.
What models does Glean use?
Glean's Model Hub provides a curated, swappable set of models from OpenAI, Google, Anthropic, Amazon, and Meta, selectable per agent and per workflow step.
How is Glean different from Agentforce?
Agentforce anchors agents on Salesforce CRM data and workflows; Glean anchors agents on a horizontal, permissions-aware index of all company knowledge, independent of any single system of record.
Executive Summary
Glean is an enterprise "Work AI" platform founded in 2019 by ex-Google search engineer and Rubrik co-founder Arvind Jain.[1] Its foundation is a permissions-aware index of company knowledge — 27B+ documents across 100+ connectors — on which it layers an AI assistant and, since early 2025, Glean Agents: a no-code builder and shared Agent Library for deploying team-wide agents that reason, plan, and take action in enterprise systems.[2][3]
The traction is among the strongest in the category. Glean surpassed $200M ARR in December 2025, doubling in nine months from the $100M mark, with $1M+ contracts growing nearly threefold; agents have executed 250M+ actions on the platform.[2] The company raised a $150M Series F at a $7.2B valuation in June 2025, led by Wellington Management, stating a target of one billion agent actions per year.[4]
| Attribute | Value |
|---|---|
| Company | Glean Technologies, Inc. |
| Founded | 2019, by Arvind Jain (CEO), T.R. Vishwanath, Piyush Prahladka, and Tony Gentilcore[1] |
| Funding | $150M Series F at $7.2B valuation (June 2025), led by Wellington Management[4] |
| ARR | $200M+ as of December 2025, doubled in nine months[2] |
| Headquarters | Palo Alto, CA[1] |
| Open Source | No — closed source, proprietary SaaS |
Product Overview
Glean's pitch is that useful enterprise agents require grounded enterprise context. The platform first builds a knowledge graph from a company's applications — documents, tickets, messages, code, CRM records — with source-system permissions enforced at query time, then exposes that context to an assistant and to agents.[3] Glean Agents, launched in early 2025 and upgraded with Agentic Engine 2 in late 2025, adds reasoning, event triggers, agent-to-agent task routing, and action execution inside connected apps.[2][3]
Key Capabilities
| Capability | Description |
|---|---|
| Agent Builder | No-code/low-code environment to build, test, and iterate on reasoning-based agents and multi-step workflows grounded in enterprise context[3] |
| Agent Library | Shared, company-wide library to discover, deploy, and reuse agents across departments[3] |
| Agentic Engine | Reasoning and orchestration layer — agents plan next steps, trigger from events, and route tasks between agents[3] |
| Knowledge graph | 27B+ indexed documents across 100+ app connectors with permissions-aware retrieval[2] |
| Model Hub | Curated multi-model access (OpenAI, Google, Anthropic, Amazon, Meta), selectable per agent and per workflow step[5] |
| Governance | Built-in security, permissions-aware governance, and real-time oversight of agent activity[3] |
Product Surfaces
| Surface | Description | Availability |
|---|---|---|
| Web app | Primary assistant, agent builder, and Agent Library | GA |
| Slack / Microsoft Teams | Assistant and agents invoked where teams already chat[3] | GA |
| In-app integrations | Zoom, Zendesk, GitHub, Miro, and other connected tools[3] | GA |
| Browser extension | Search and assistant alongside any web app[3] | GA |
| API | Programmatic access to search, assistant, and agents | GA |
Technical Architecture
Glean is a managed, closed-source SaaS platform. The architecture stack is: connectors ingest content from 100+ enterprise apps into a permissions-aware index and knowledge graph; the Agentic Engine performs retrieval, reasoning, and orchestration on top; and the Model Hub abstracts the LLM layer so customers can mix models from multiple providers without re-engineering agents.[2][6]
Key Technical Details
| Aspect | Detail |
|---|---|
| Deployment | Managed SaaS (SaaS-hosted or customer cloud-hosted options)[7] |
| Model(s) | Multi-model via Model Hub — OpenAI, Google, Anthropic, Amazon, Meta; per-agent and per-step selection[5] |
| Integrations | 100+ connectors (Google Workspace, Microsoft 365, Slack, Jira, Confluence, Salesforce, Zendesk, GitHub, and more)[2] |
| Open Source | No |
Strengths
- Verified revenue traction — $200M ARR as of December 2025, doubled in nine months, with $1M+ contracts up nearly 3x; one of the few agent platforms with disclosed nine-figure revenue[2]
- Grounding moat — agents run over a permissions-aware index of 27B+ documents and 100+ connectors, which directly addresses the hallucination and access-control problems that sink generic agent deployments[2]
- Agents in production volume — 250M+ agent actions executed, up from 100M+ annually at the June 2025 Series F, with a stated 1B/year target[2][4]
- Model-agnostic by design — the Model Hub avoids single-LLM-vendor lock-in and lets builders route each workflow step to the cheapest adequate model[6]
- Meets teams where they work — Slack, Microsoft Teams, Zoom, Zendesk, GitHub, browser extension, and API surfaces[3]
- Well capitalized — $150M Series F at $7.2B with Wellington, Sequoia, Lightspeed, Kleiner Perkins, and General Catalyst on the cap table[4]
Cautions
- Agents are one pillar, not the whole product — Glean is first an enterprise search/assistant suite; teams evaluating it purely as an agent platform are buying (and paying for) the full Work AI stack. The category overlap with enterprise search makes apples-to-apples comparison against agent-first platforms imperfect.
- No public pricing — enterprise sales only; third-party analyses estimate roughly $35–50/user/month with ~100-seat minimums and fully-loaded mid-to-large deployments in the hundreds of thousands of dollars per year[7]
- Closed source — no self-hosted open-source option and no way to inspect or extend the core platform
- Indexing prerequisite — value depends on connecting and indexing company systems first; rollout is an IT project, not a self-serve signup
- Big-platform squeeze — Microsoft Copilot, Google Gemini Enterprise, and ChatGPT Enterprise bundle adjacent capability into suites enterprises already pay for
- Per-seat economics for agents — seat-based pricing fits assistants better than asynchronous background agents, where consumption models (e.g., Agentforce's per-action Flex Credits) map more directly to value
Pricing & Licensing
| Tier | Price | Includes |
|---|---|---|
| Enterprise | Custom (not publicly listed) | Search, assistant, Glean Agents, Model Hub, governance; quoted per user with volume discounts[7] |
Third-party estimates put per-user costs around $40/user/month SaaS-hosted or $35/user/month customer-hosted, with ~100-seat minimum commitments and add-ons for advanced generative capability.[7]
Licensing model: Proprietary closed-source SaaS, annual enterprise contracts.
Hidden costs: Implementation and connector rollout, admin/governance staffing, and seat minimums; analyses peg fully-loaded annual spend for mid-to-large organizations at $350K–$480K.[7]
Competitive Positioning
Direct Competitors
| Competitor | Differentiation |
|---|---|
| Agentforce | Agentforce is CRM-native with consumption pricing; Glean is horizontal, grounding agents in all company knowledge rather than Salesforce data |
| Dust | Dust is a lighter-weight, self-serve team agent builder; Glean pairs agents with a deep enterprise search index and sells top-down to large enterprises |
| Microsoft Copilot / Copilot Studio | Copilot is Microsoft 365-native and bundled; Glean is cross-suite, indexing Google Workspace, Atlassian, Salesforce, and Microsoft alike |
| Google Gemini Enterprise | Gemini Enterprise anchors on Workspace and Google Cloud; Glean is vendor-neutral across both Google and Microsoft estates |
| ChatGPT Enterprise | ChatGPT Enterprise leads with the model; Glean leads with the permissions-aware knowledge graph and multi-model choice[6] |
When to Choose Glean Over Alternatives
- Choose Glean when: you want shared team agents grounded in all company knowledge with enforced permissions, and you span multiple suites (Google + Microsoft + Atlassian + Salesforce)
- Choose Agentforce when: your workflows live in Salesforce and you want consumption-priced customer-facing agents
- Choose Dust when: you want fast, self-serve agent building for a smaller team without an enterprise sales cycle
- Choose Copilot/Gemini when: you are single-suite and want bundled pricing over best-of-breed
Ideal Customer Profile
Best fit:
- Enterprises (hundreds to tens of thousands of seats) with knowledge sprawled across many SaaS tools
- Organizations where permissions-correct retrieval is a hard requirement (regulated industries, large orgs)
- Teams that want a shared Agent Library non-engineers can build for and deploy from
- Buyers who want multi-model flexibility rather than a single LLM vendor
Poor fit:
- Small teams or startups — seat minimums and enterprise sales cycles price them out[7]
- Open-source or self-host requirements
- Developer teams wanting a code-first agent framework
- Single-suite shops content with bundled Copilot or Gemini
Viability Assessment
| Factor | Assessment |
|---|---|
| Financial Health | Strong — $200M ARR doubling in nine months; $150M Series F at $7.2B with blue-chip investors[2][4] |
| Market Position | Leader in horizontal Work AI — Gartner Tech Innovator in Agentic AI; 2025 CNBC Disruptor 50[2][4] |
| Innovation Pace | Rapid — agents launched early 2025, Agentic Engine 2 and third-gen assistant by late 2025[2] |
| Community/Ecosystem | Enterprise-centric — 100+ connectors, no open-source community[2] |
| Long-term Outlook | Positive, with platform-bundling risk from Microsoft, Google, and OpenAI |
Glean has converted enterprise search into an agent platform with real revenue, and the grounding/permissions layer is a genuine moat against model-first competitors. The open question is whether a best-of-breed horizontal platform can hold pricing power as suite vendors bundle "good enough" agents into existing contracts.
Bottom Line
Glean is the strongest horizontal entrant in team agent platforms: $200M ARR, 250M+ agent actions, and a defensible answer to the question every enterprise agent deployment hits first — what can this agent see, and is it allowed to?[2] Its agents inherit a permissions-aware index that agent-first startups must rebuild from scratch. The trade-offs are equally clear: closed source, opaque enterprise-only pricing, and the fact that agents ride along with a broader search/assistant suite you must buy whole.
Recommended for: Multi-suite enterprises that want governed, shared AI agents grounded in all company knowledge, accessible from Slack, Teams, and the web.
Not recommended for: Small teams, open-source shops, code-first agent builders, or single-suite organizations satisfied with bundled Copilot/Gemini.
Outlook: Strong. Doubling to $200M ARR in nine months while tripling $1M+ contracts shows the top-down motion is working, and the 1B agent-actions target signals where the product is headed.[2][4] The main risks are suite-vendor bundling and per-seat pricing friction as agent work shifts from assistants to autonomous background execution.
Research by Ry Walker Research • methodology
Sources
- [1] Glean Revenue, Funding & News (Sacra)
- [2] Glean Surpasses $200M in ARR for Enterprise AI (Glean Press)
- [3] Glean Agents Product Page
- [4] Glean Raises $150M Series F at $7.2B Valuation (Glean Press)
- [5] Set up LLMs using the Model Hub (Glean Docs)
- [6] Glean Model Hub Product Page
- [7] Glean Pricing: Costs, Hidden Fees & TCO Breakdown (Workativ)