KIDZ AI has entered into a five-year GPU compute services agreement with Canopy Wave valued at approximately $44.6 million. The contract supports KIDZ AI’s expansion beyond education technology into AI computing infrastructure and GPU cloud services.
Under the agreement, Canopy Wave is expected to pay KIDZ AI for access to a dedicated cluster of 256 NVIDIA Blackwell B300 GPUs. The arrangement remains subject to KIDZ AI’s subsidiary placing a non-cancellable order for the servers required to deliver the contracted services.
KIDZ AI’s wholly owned subsidiary, Catalyst Compute, plans to deploy the GPUs across 32 specialized computing nodes. Each node is expected to include dual Intel Xeon 6776P processors, four terabytes of DDR5 memory and 800-gigabit-per-second InfiniBand networking.
InfiniBand provides high-speed connections between servers and GPUs, helping large AI workloads exchange data with minimal delay. This infrastructure is particularly important for running advanced models that require multiple processors to work together.
Canopy Wave operates an AI inference platform designed for high-throughput and cost-efficient deployment of open-weight models. Its supported model families include technologies from Moonshot AI, DeepSeek, Qwen and Zhipu AI.
AI inference is the process of running a trained model to generate predictions, responses or other outputs. Unlike model training, which involves teaching a system using large datasets, inference represents the ongoing computing activity required when customers use an AI application.
Open-weight models make their trained parameters available for organizations to deploy and customize. This can give enterprises greater control over infrastructure, privacy and operating costs than services that are accessible only through proprietary external platforms.
KIDZ AI is positioning the contract as an example of its differentiated neocloud strategy. Neocloud providers offer specialized computing infrastructure optimized for AI workloads, typically using large clusters of advanced GPUs.
Rather than installing GPU capacity before securing customers, KIDZ AI intends to pair infrastructure investments with long-term contracted demand. The company believes this model can reduce the financial risk associated with building speculative computing capacity.
Canopy Wave plans to use the cluster to provide secure, enterprise-grade inference services. The companies expect the infrastructure to support workloads requiring high throughput, low latency and strong data privacy controls.
The agreement reflects KIDZ AI’s view that economic efficiency will become increasingly important as AI adoption expands. Competitive factors may include the cost of producing each model output, GPU utilization, processing speed and the system’s ability to serve many users simultaneously.
KIDZ AI believes improving these economics could make advanced AI available to a broader range of enterprises. The company sees demand for infrastructure capable of running increasingly powerful open-weight models at lower costs.
The contract also represents a significant step in KIDZ AI’s transformation into a broader AI technology and infrastructure business. Formerly known as Classover Holdings, the company continues to develop AI-powered education products while expanding into GPU platforms, data centers and related infrastructure.
The companies said the deployment could serve as a model for additional contracted GPU clusters. However, KIDZ AI’s ability to realize the full value of the agreement depends on ordering, financing and successfully deploying the required equipment.
KEY QUOTES:
“Today’s agreement is just the beginning of our journey with KIDZ AI. We look forward to helping KIDZ AI scale additional GPU clusters and turn this infrastructure into powerful token factories for the enterprise AI market.”
“By combining KIDZ AI’s next-generation compute platform with Canopy Wave’s optimized inference capabilities, we believe we can deliver secure, cost-efficient, enterprise-grade AI inference services with strong data privacy protections to enterprises across the United States.”
Tao Zhang, CEO of Canopy Wave
“This milestone contract is an important step in KIDZ AI’s evolution into a broader AI technology and infrastructure platform.”
“What we believe sets our neocloud model apart is pairing next-generation computing infrastructure with long-term contracted enterprise demand, a scalable commercial foundation rather than speculative capacity, and a model we aim to repeat across future deployments and partnerships.”
“Cost-efficient open-weight models such as DeepSeek and Kimi are shifting the AI market from a race defined by raw scale to one increasingly defined by economic efficiency.”
“Canopy Wave’s strength in optimized inference aligns directly with this shift, and we believe affordable AI will be one of the most powerful structural forces driving the next phase of adoption.”
Stephanie Luo, CEO of KIDZ AI
