Ghost Raised $11 Million for Personal AI Computer
The San Francisco-based startup plans to ship a local-inference machine designed to run AI models offline by October 2026.
Updated on Oct. 5, 2026 in Artificial Intelligence
Live Poll
Would you pay for personal AI hardware to keep your data off the cloud?
San Francisco startup Ghost has emerged from stealth with $11 million in seed funding led by Andreessen Horowitz. The company announced the Core, a dedicated hardware device designed for local AI processing.
Why it matters
The company aims to enable personal ownership of AI models and user data by moving inference away from cloud infrastructure. By keeping processes local, the hardware allows users to maintain complete control over their context and model interactions.
The Core computer is rated at 770 AI TOPS and powered by an NVIDIA RTX PRO 4000 Blackwell GPU. It is equipped with 24 GB of GDDR7 memory, 64 GB of DDR5 RAM, a 1 TB NVMe SSD, and a 6-core AMD Ryzen 5 7600 processor.
The players
Ghost
A San Francisco-based startup building dedicated hardware for local AI model inference.
Andreessen Horowitz
A Silicon Valley venture capital firm that invests in software, infrastructure, and artificial intelligence companies.
Zain Javaid
The CEO of Ghost.
The details
The machine runs inference locally, meaning the system computes model responses directly on the hardware rather than sending data to external servers. This is achieved through a stack that connects with local apps and devices, utilizing models such as Qwen 3.8-27B and Gemma 4-31B. By executing these workloads offline, the device facilitates local interaction without reliance on third-party cloud infrastructure.
Timeline
October 5, 2026: Ghost emerged from stealth.
October 31, 2026: Shipments of the Core computer are expected to begin.
The Tech Race
Ghost enters the market for high-performance local computing, positioning the Core as a specialized alternative to general-purpose workstations. The firm aims to capture user segments moving away from cloud-dependent AI tools toward local ownership.
The Core computer is priced at $3,499 and will be available to consumers who prioritize offline model access. Prospective users should note that the device requires hardware-level integration with local apps and remains unavailable until shipments commence in late 2026.
The takeaway
Ghost is betting that high-end local hardware will become the standard for privacy-focused AI users. Watch for the October 31, 2026 shipping milestone to see how the system performs against cloud-based alternatives in real-world benchmarks.
What happens next
Shipments of the Core computer are scheduled to begin on October 31, 2026.
Further reading
For more on how hardware is adapting to run LLMs, visit the Artificial Intelligence section.
Live Poll
Would you pay for personal AI hardware to keep your data off the cloud?
