Hark Secures $700M Series A for Universal AI Interface

AI laboratory Hark has secured a $700 million Series A funding round at a $6 billion post-money valuation. The investment, led by Parkway Venture Capital, funds the development of a vertically integrated universal AI interface. Founded by Brett Adcock, the company is building proactive systems designed to operate beyond traditional chat windows by integrating foundation models, software, and dedicated hardware.

Vertically Integrated Architecture

Hark’s technical strategy relies on owning the entire compute stack. The company is currently training its next-generation models on a data center powered by NVIDIA B200 GPUs. The initial software release, scheduled for Summer 2026, will feature multimodal foundation models built around persistent memory. These models natively process speech, text, and vision inputs to function as a unified personal platform. For developers evaluating agent architectures, Hark’s design signals a shift away from stateless API calls toward continuous, stateful execution.

Purpose-Built Hardware

Software represents only half of the company’s roadmap. Following the Summer 2026 model release, Hark plans to introduce a family of dedicated AI devices. These devices target individual and home environments, deliberately diverging from the form factors of existing smartphones, smart glasses, or wearables. The hardware division is led by Head of Design Abidur Chowdhury, who previously directed industrial design for several flagship Apple products. The $700 million capital injection will primarily fund scaling these hardware and AI research teams beyond their current headcount of 70 employees.

Strategic Market Positioning

The investor syndicate highlights the industry’s interest in native hardware integration. The round includes participation from the venture arms of nearly every major silicon provider, including NVIDIA, AMD Ventures, Intel Capital, and Qualcomm Ventures. Salesforce Ventures, ARK Invest, Brookfield, Greycroft, Prime Movers Lab, and Tamarack Global also participated. Hark’s leadership points to an everyday consumer focus as their primary differentiator. While competitors have increasingly optimized their models for developer tooling, Hark is engineering systems for ambient consumer deployment.

If you build applications relying on large language models, the introduction of consumer hardware with persistent ambient context changes the deployment landscape. You should evaluate agent memory strategies within your own stack, as users will increasingly expect systems that maintain state across sessions and modalities natively.