Hark Secures $700 Million Series A Funding at $6 Billion Valuation

On May 21, 2026, AI hardware startup Hark announced the completion of a $700 million Series A funding round, valuing the company at $6 billion. The round was led by Parkway Venture Capital, with participation from Nvidia, AMD Ventures, ARK Invest, Intel Capital, Qualcomm Ventures, and Salesforce Ventures.

Single-round funding scale sets AI startup record

Securing $700 million at the Series A stage is relatively rare in the AI startup space. Hark, founded only about six months ago, has attracted simultaneous investments from multiple chip and software giants, reflecting investor recognition of its hardware-model integration approach. The company has already secured NVIDIA B200 data center resources for model training, providing practical computing power support for subsequent product deployment.

Founder's serial entrepreneurship background earns trust

Hark founder Brett Adcock previously founded Figure.AI and Archer. Figure.AI focuses on humanoid robots, while Archer is advancing commercialization in the electric vertical takeoff and landing aircraft space. Both companies have entered product validation stages. Adcock personally invested $100 million in Hark, further strengthening investors' confidence in the team's execution capability.

Model and hardware integration path

Hark positions itself at the intersection of foundational models for AI personal assistants and dedicated hardware devices, aiming to build a universal AI interaction interface. The company plans to first release a batch of AI models this summer, followed by the launch of AI-native hardware products. Compared to a pure software approach, this integration method requires simultaneously solving three challenges: model training, hardware adaptation, and user interaction. Currently, the company has approximately 70 employees, with a focus on actual model operation and hardware prototype development.

The company emphasizes that its initial models will run on real training data, rather than remaining at the demonstration stage.

Next product plans

• Release the first batch of AI models in summer 2026, supporting stable API calls.
• Subsequently launch dedicated AI hardware devices, aiming to lower the barrier for user adoption.
• Continuously expand the NVIDIA B200 cluster scale to ensure manageable training and inference costs.

Based on disclosed information, Hark has completed team formation, computing power procurement, and multiple investment rounds in its early stages, demonstrating a strong execution pace. Whether subsequent products are delivered on time and whether actual performance aligns with promotional claims will be key areas for market scrutiny.