Mount Thor, a San Francisco company founded in 2026, announced last week that it has raised $45 million in equity capital to build cloud infrastructure on Apple hardware for AI agents and related workloads. The total spans Seed and Series A financings, with CRV and Conviction co-leading the Series A. The company develops, deploys and manages rack-scale systems of Apple hardware, running macOS on Apple silicon as managed infrastructure for AI research and deployment. No valuation was disclosed.
Who put in the money
CRV and Conviction co-led the Series A. Offline Ventures co-led the Seed. The split between the two rounds was not disclosed, so the $45 million figure covers both financings together.
CSC Leasing, Enea Capital, Long Journey Ventures and Upper90 participated, alongside strategic angels from Cognition, Core Automation, Daytona, Meta, OpenAI, Together AI and Vercel. Those angels invested as individuals; the companies they work for are not themselves listed as investors in the round.
Separately from the equity, Mount Thor secured non-dilutive financing to support hardware purchases. The amount was not disclosed, and it sits outside the $45 million equity total.
What Mount Thor is building
The pitch is Apple hardware as elastic compute. Per 9to5Mac, Mount Thor runs macOS on Apple silicon as elastic compute, and managed compute can be provisioned either as a full bare-metal Mac or as an isolated macOS virtual machine, through a web portal, a command-line tool, Kubernetes, or the API.
The company says its infrastructure is meant to serve five use cases: research labs building diverse, high-quality reinforcement learning environments for frontier computer-use models; agent infrastructure providers offering secure macOS sandboxes to test new capabilities; agentic products running macOS instances for weeks at a time, so agents can work across a user’s different devices under the same Apple identity; AI-native software development platforms building, testing and simulating software for Apple’s devices; and enterprises reducing inference costs by moving workloads to open-source models optimized for Apple silicon.
Mount Thor says it deployed its first data center in less than six months and plans to expand rapidly to meet customer demand. The team previously built infrastructure at Cerebras, CoreWeave, Fluidstack, Nebius, OpenAI and xAI. The new capital goes toward expanding the compute footprint and growing the team, with a stated focus on standing up an inference lab for open-source models on Apple silicon.
What the raise means for vendors
A company that just closed $45 million and is buying Macs at rack scale is an active buyer on several fronts at once: data center space and power, networking and storage, fleet management and device tooling, security and compliance review, and recruiting for infrastructure and inference engineering roles. The non-dilutive hardware financing is a reminder that the cost base here is physical fleets, not only software. Vendors tracking AI startups and San Francisco AI startups should treat the first-data-center milestone as the start of a procurement cycle, not the end of one.
The round also sits inside a broader funding pattern around agent plumbing. LangChain’s $300 million Series C also landed last week, on the same mid-November stretch of agent-infrastructure funding. For anyone selling into cloud computing startups, the macOS-specific slice of that market is newly funded. More names from the same week are in our list of recently funded startups.
Free: top-10 + totals • Paid: full list + verified contacts
This is free coverage of a funding event. Growth and Scale unlock every newly funded company we track each week, plus verified decision-maker contacts.
Sources: Mount Thor via Business Wire, 12 November 2026; Mount Thor; 9to5Mac. Featured image: San Francisco from Twin Peaks, 2022, by Dietmar Rabich, Wikimedia Commons, CC BY-SA 4.0.