Mount Thor, a San Francisco startup that rents out Apple hardware at data-center scale, has raised $45 million to build Mac data centers for AI agents. The money came across Seed and Series A rounds, with CRV and Conviction co-leading the Series A. The company’s bet is simple: AI agents increasingly need to use real computers, and many of the tasks they handle only run on macOS.
Inside Mount Thor’s $45M Funding Round
The Series A was co-led by CRV and Conviction, while Chad Byers and Offline Ventures co-led the earlier Seed round. Other backers include CSC Leasing, Enea Capital, Long Journey Ventures and Upper90. Angel investors from OpenAI, Meta, Cognition, Together AI, Vercel, Daytona and Core Automation also joined.
Mount Thor says it also raised separate non-dilutive capital to help pay for hardware. The company was founded in 2026 and deployed its first data center in less than six months. Its team previously built infrastructure at Cerebras, CoreWeave, Fluidstack, Nebius, OpenAI and xAI.
Funding Details at a Glance
| Item | Detail |
|---|---|
| Total raised | $45 million in equity (Seed and Series A) |
| Series A leads | CRV and Conviction |
| Seed leads | Chad Byers and Offline Ventures |
| Extra financing | Non-dilutive capital for hardware purchases |
| Founded | 2026, headquartered in San Francisco |
| Main uses | Agent sandboxes, computer-use training, app testing, inference |
Why AI Agents Need Mac Data Centers
A growing class of AI agents works by operating a computer the way a person would, clicking, typing and moving between apps. Many consumers and businesses use Macs, and some tools only exist on macOS, so agents need to learn and work in that environment too. Mount Thor argues that demand for Apple hardware will now grow with the number of agents, not just the number of people.
Founder and CEO Peter Voell called Apple’s platform at data-center scale “a massive untapped opportunity in AI infrastructure.” He also pointed to macOS accessibility features as a strong base for agents to work while humans keep them accountable. If you have tried AI Mac automation, you have seen a small version of this idea.
What Customers Can Do With It
Mount Thor runs macOS on Apple silicon as rentable cloud instances, much like providers rent Linux virtual machines. The company lists several target uses:
- Building reinforcement learning environments to train computer-use models.
- Running secure macOS sandboxes where agents can test new skills.
- Keeping macOS instances alive for weeks so an agent can work across a user’s devices under one Apple ID.
- Building and testing software for Apple devices at scale.
- Moving inference to open-source models running on Apple silicon to cut costs.
Customers can get either a full bare-metal Mac or an isolated macOS virtual machine with admin access, billed per Mac on a 24-hour basis. Larger organizations can also place their own Macs in a Mount Thor facility and manage them with their own tools.
Mac Data Centers as an Inference Platform
The other half of the pitch is running AI models. The M5 Ultra Mac Studio, which shipped in August, offers up to 512GB of unified memory, enough to fit large open-weight models on one machine. Mount Thor says a key goal for the new money is building a top inference lab for open-source models on Apple silicon.
Renting Macs in the cloud is not new, since AWS and MacStadium already do it. The difference here is the customer: Mount Thor is aiming at AI agents and labs rather than human developers. That fits a wider trend toward agent sandboxes and local AI agents running on dedicated hardware.
What This Means for You
Most people will never rent a Mac fleet themselves, but this kind of infrastructure shapes the AI tools you use. Here is what to take away:
- AI agents that can operate Mac apps for you may get better faster, since labs can now train them on many macOS machines at once.
- Developers building Apple apps could see quicker, cheaper testing from AI-powered build farms.
- Businesses running open models may find Apple silicon a lower-power option to GPUs for some inference jobs.
- Security matters here: as agents get their own computers, keep an eye on agent security tools and permissions.
The next big user of the Mac might not be a person at all, but an AI agent working on someone’s behalf.
Sources
Primary source: Mount Thor’s funding announcement. Additional reporting from 9to5Mac.
Conclusion
Mount Thor’s $45 million raise is a bet that AI agents will become major users of Apple hardware. By renting macOS and Apple silicon at data-center scale, the startup wants to serve labs training computer-use models, agent platforms and companies chasing cheaper inference. Backing from CRV, Conviction and angels at top AI companies gives it room to grow fast. Whether Macs become a real pillar of AI infrastructure will depend on how quickly agent demand actually scales.
