Granola has secured $125 million in fresh capital at a $1.5 billion valuation, positioning the company to evolve from a popular meeting notetaker into a broader enterprise AI application. The Series C was led by Danny Rimer at Index Ventures with participation from Mamoon Hamid at Kleiner Perkins, alongside existing backers Lightspeed, Spark, and NFDG. Less than a year after closing $43 million, Granola’s total funding now stands at $192 million.
The pitch is straightforward: people may dislike an obvious bot in the call, but they rarely object to a quiet desktop app that captures high-quality notes. That low-friction entry has given Granola distribution and data context—two ingredients it now aims to weaponize for enterprise workflows.
From Notetaker to Full Enterprise Platform
Granola began as a prosumer desktop assistant that transcribes meetings and auto-generates clean, shareable notes. Over the past year, it layered collaboration features so teammates could co-edit and refine action items. The company says it has since made inroads at fast-scaling companies including Vanta, Gusto, Thumbtack, Asana, Cursor, Lovable, Decagon, and Mistral AI.
With the new round, Granola is rolling out Spaces—team workspaces with Folders and granular permissions to control who can see which conversation, summary, or action item. Users can query within a Space or folder to cut through noise and surface the exact snippet, decision, or follow-up they need.
The move reflects a broader shift: meeting notes are table stakes. The enterprise win lies in turning that raw signal into structured knowledge and downstream actions that plug into the tools where work actually gets done.
APIs and MCP Turn Meeting Notes into Enterprise Actions
After introducing a Model Context Protocol (MCP) server in February—part of an emerging standard championed by Anthropic to let AI agents securely access tools and data—Granola is launching two APIs. A personal API gives users programmatic access to their own notes and those shared with them (available on business and enterprise plans). An enterprise API lets admins work with team-level context (enterprise only).
Granola is also updating its MCP server so agents can navigate notes within folders and shared spaces. The app already plugs into popular AI and developer tools including Claude, ChatGPT, Lovable, Figma Make, Replit, Manus, v0, Bolt.new, Duckbill, and Dreamer, with more integrations in the works.
In practical terms, that means a sales call summary can automatically enrich a CRM record; engineering discussions can spawn Jira issues with clear acceptance criteria; and customer feedback collected across multiple meetings can be clustered, prioritized, and routed to the right owners—without manual copy-paste.
Course Correction After Developer Backlash
The API push follows a developer dust-up earlier this year when Granola changed how it stored local data, breaking some on-device agent workflows. Power users—including a partner at Andreessen Horowitz—publicly criticized the move. Co-founder Chris Pedregal said the company had not intended to “lock down” data but that the previous local cache wasn’t designed for AI agents at scale. He pledged bulk-access capabilities and better agent support; today’s APIs and MCP updates make good on that promise.
For enterprise buyers, that response matters. Openness and predictable data access patterns are quickly becoming procurement criteria, not nice-to-haves, as AI agents move from experiments to production systems.
Why Investors Are Leaning In on Granola’s Strategy
Investors have seen this movie: tools that start with a compelling single-player utility often earn the right to become a system of engagement. The flywheel here is attractive—ubiquitous note capture yields high-quality context, and context is what makes AI outputs accurate, actionable, and auditable.
Market timing also helps. McKinsey estimates generative AI could create $2.6–$4.4 trillion in annual economic value as companies bake it into sales, software engineering, and customer operations. Meeting assistants are among the first sanctioned AI deployments inside many organizations because they deliver clear, measurable savings in time-to-summary and follow-up execution.
A Crowded Field and the Enduring Moat Question
Competition is fierce. Read AI, Fireflies, Quill, and others are pushing beyond transcription toward action-oriented workflows. Differentiation will hinge on three things: depth of enterprise integration (CRMs, help desks, code repos), trustworthy retrieval across private corpora, and security controls that satisfy legal and compliance teams.
Accuracy and governance will be scrutinized. Enterprises will expect transparent provenance—what came from the transcript, what the model inferred, and where the data lives—as well as granular retention and role-based access. Vendors that can prove lower false-positive rates in task extraction and faster time-to-action will have an edge.
Granola’s bet is that Spaces, robust APIs, and an MCP-first posture will let it own the “context layer” across meetings, chats, and docs. If it succeeds, the meeting is just the ingestion point; the real product is the mesh of structured knowledge that fuels agents across the stack.
For now, a $125 million round at a $1.5 billion valuation sets the stage—and the bar. The next chapter will be written not by prettier summaries, but by how reliably those summaries trigger the right actions in the rest of the enterprise.
