XMax claims $25m AI API contracts and enters GPU-as-a-Service
XMax Inc., a Nasdaq-listed company that rebranded from furniture retailer Nova LifeStyle and is now pursuing an artificial intelligence pivot, has announced that it has executed multiple enterprise AI model API service agreements with a combined contractual value of up to approximately $25 million. The contracts are usage-based, meaning the stated figure represents a ceiling rather than guaranteed revenue. In parallel, the company says it has signed its first commercial GPU-as-a-Service (GPUaaS) agreement, marking its formal entry into the managed compute market.
The agreements are described as covering enterprise AI model API access, cloud infrastructure integration, and value-added platform services. XMax says customers are seeking scalable inference capabilities, secure API connectivity, and usage-based commercial models. No customer names, contract durations, or minimum-commitment figures were disclosed in the release.
The pivot in context
XMax's background is worth noting for readers assessing the announcement. The company, headquartered in Commerce, California, operates an established furniture design and distribution business alongside its AI subsidiary, XMax AI Inc. The AI unit is a relatively recent addition, and the company has not disclosed what proportion of total group revenue the technology business currently generates. Chief executive Mr Lu said in a statement that the contracts and the GPUaaS entry "represent meaningful progress" in executing the company's AI growth strategy, though the release provides no performance benchmarks, named partners, or reference customers to substantiate that claim.
The $25 million aggregate figure is subject to "actual service usage and consumption levels," a standard qualifier in consumption-based contracts that can result in substantially lower realised revenue. Enterprise buyers in AI infrastructure often negotiate framework agreements with high notional ceilings but modest initial commitments.
Market landscape
The GPUaaS and AI API brokerage segments are increasingly crowded. Hyperscalers including AWS, Microsoft Azure and Google Cloud dominate enterprise AI infrastructure, while a cohort of specialist GPU cloud providers such as CoreWeave, Lambda Labs and Voltage Park has attracted significant venture capital over the past two years. A number of smaller Nasdaq-listed technology companies have also announced AI pivots, a trend regulators and investors have scrutinised given the distance between press-release announcements and recognised revenue.
The managed inference layer is similarly competitive, with vendors ranging from large cloud providers to emerging model-serving startups competing on latency, pricing transparency, and compliance posture. For a company with XMax's profile to win durable enterprise contracts in this environment, buyers will want evidence of uptime guarantees, security certifications such as SOC 2 Type II or ISO 27001, and demonstrated throughput benchmarks.
What comes next
XMax says it intends to pursue additional enterprise customers, strategic partnerships, and international market opportunities, though it offered no timeline or geographic specifics. Investors will be watching for contract-to-revenue conversion: the gap between a signed usage-based framework and billable consumption is the key risk in this commercial model. A follow-on disclosure naming at least one anchor customer and publishing utilisation data would materially strengthen the company's stated narrative of building an integrated AI computing ecosystem.