SAIHEAT and Canopy Wave combine to form open-weight AI inference operator

SAIHEAT Limited is merging with Santa Clara-based Canopy Wave in an equity-based transaction that creates a U.S.-headquartered provider of AI inference infrastructure. The deal shifts the combined entity's primary focus toward hosting open-weight large language models while retaining modular data center operations.

In a strategic repositioning designed to capitalize on expanding enterprise workloads for open-weight artificial intelligence models, SAIHEAT Limited has signed a definitive business combination agreement with Canopy Wave, Inc. Under the transaction framework, Canopy Wave will become a wholly owned operating subsidiary of the company, which will be rebranded as Canopy Wave Holdings Inc. and trade on the Nasdaq Stock Market under the symbol CWAV. Following the completion of the merger, expected before the end of the year, leadership will transition to Canopy Wave's founding executives, and operational headquarters will move to Santa Clara, California.

SAIHEAT operates modular computing power facilities designed to optimize energy utilization near power sources, while Canopy Wave functions as an artificial intelligence infrastructure developer specializing in orchestration software, API endpoints, and cloud GPU deployment. Upon closing, former Canopy Wave equity holders will maintain approximately 54.19 percent of the combined corporate equity and 78.44 percent of total voting interest, supported by a concurrent private investment in public equity of Class A ordinary shares valued at approximately 4.5 million USD at a price of 18.15 USD per share.

The strategic shift reflects a broader transition across the global technology sector, where enterprise spending is increasingly migrating from capital-intensive initial model training toward ongoing inference deployment. As open-weight artificial intelligence models achieve capability parity with proprietary equivalents, business adoption is driving sustained demand for secure, cost-effective API services, automated coding support, and autonomous agent workflows. Integrating physical data center infrastructure with dedicated software management allows the unified platform to optimize token generation costs and address data privacy concerns through enterprise security standards.

For technology investors and corporate IT decision-makers, the merger illustrates growing commercial momentum around open-source artificial intelligence deployment over proprietary platforms. By securing direct access to public capital markets and combining data center operations with high-performance GPU leasing arrangements, the transaction establishes a scalable infrastructure footprint. This model addresses expanding demand across enterprise software developers, automated digital services, and cloud platform managers seeking specialized, high-throughput computing capabilities.