<table name="logo_release" border="0" cellspacing="10" cellpadding="5" align="right">   <tbody>    <tr>     <td><img src="https://mmx.prnasia.com/media/MS1872144/ZETTABYTE-Logo-Horizontal-WhiteBG-Logo.jpg?id=OA2865046&amp;p=medium600" border="0" alt="" title="logo" hspace="0" vspace="0" width="118" /></td>    </tr>   </tbody>  </table>  <p><i>zCLOUD delivers Model-as-a-Service (MaaS), hosting open-weight and open-source models such as Kimi, GLM, Llama, GPT-OSS, Gemma, and Qwen on Zettabyte's GPUs and making them available through a single API.</i></p>  <ul type="disc">   <li>Offers open-weight and open-source models through an API, running on Zettabyte's GPUs;</li>   <li>Provides access to NVIDIA H100, H200, B200 and B300 GPUs;</li>   <li>Offers three fully managed ways to access capacity, backed by a 99.5 percent uptime SLA target;</li>  </ul>  <p><span class="legendSpanClass">TAIPEI CITY, Taiwan</span>, <span class="legendSpanClass">Aug. 13, 2026</span> /PRNewswire/ -- Zettabyte, a global AI computing company, today announced the general availability of Model-as-a-Service on zCLOUD, its GPU cloud. Teams can efficiently call open-source models through API on Zettabyte's NVIDIA H100, H200, B200 and B300 GPUs.</p>  <p><b>Model-as-a-Service<br /></b>With Model-as-a-Service, zCLOUD hosts popular open weight and open-source models, ready to call through an API and run on Zettabyte's GPUs. Teams can build on them directly, without sourcing, installing, and tuning a model themselves. Teams that want raw capacity can still rent GPUs on zCLOUD and run their own models.</p>  <p><b>GPU Availability<br /></b>NVIDIA H100 is available today, with pricing from US$1.99 per GPU-hour depending on needs and scale. zCLOUD is backed by a 99.5 percent uptime SLA target, and reserved clusters are built to deploy in under six hours from confirmed requirements.</p>  <p><b>Ways to Access GPU Capacity<br /></b>Zettabyte offers three ways to access capacity:</p>  <ul type="disc">   <li><b>Short-Term Capacity</b>&nbsp;- GPU instances that scale to hundreds of GPUs, with no long-term commitment.</li>   <li><b>Reserved Clusters</b>&nbsp;- Predictable, fully managed capacity for sustained training and inference under a term commitment.</li>   <li><b>Private Cloud</b>&nbsp;- Custom deployments for teams with procurement, security, or region requirements.</li>  </ul>  <p>&quot;AI computing will be everywhere, from hyperscale data centers to the edge. The supporting AI infrastructure and software must be compatible, resilient, and agile. zCLOUD was built to deliver Model-as-a-Service, bringing more than 40 hardware providers together and managing the underlying AI infrastructure through our zSUITE full stack,&quot; said Dr. David Ku, Technology Leader at Zettabyte, former Corporate Vice President and CTO of AI at Microsoft.</p>  <p>Model-as-a-Service on zCLOUD is generally available today at <a href="https://edge.prnewswire.com/c/link/?t=0&amp;l=en&amp;o=4751133-1&amp;h=1762725254&amp;u=https%3A%2F%2Fwww.zettabytecloud.com%2F&amp;a=www.zettabytecloud.com" target="_blank" rel="nofollow" style="color: #0000FF">www.zettabytecloud.com</a>.</p>  <p><b>About Zettabyte<br /></b>Zettabyte is a global AI computing company focused on delivering and measuring high-quality compute. Through its zSUITE platform, the company enables organizations to improve reliability, readiness, observability, cost transparency, and energy efficiency across AI infrastructure environments. Learn more at <a href="https://edge.prnewswire.com/c/link/?t=0&amp;l=en&amp;o=4751133-1&amp;h=3385331479&amp;u=https%3A%2F%2Fwww.zettabyte.space%2F&amp;a=www.zettabyte.space" target="_blank" rel="nofollow" style="color: #0000FF">www.zettabyte.space</a>.</p>