NexGen Cloud, a leading provider of enterprise AI cloud solutions has launched access to OpenAI’s latest open-weight GPT-OSS models – including the 120-billion and 20-billion parameter versions – through its Hyperstack platform. Within Hyperstack, the AI Studio service provides serverless inference for the 120B model, enabling instant, plug-and-play access without the need to manage infrastructure.
Only days after GPT-OSS launched, NexGen Cloud became the first UK-owned GPU cloud to make the 120B model available in this way, allowing businesses to work with OpenAI’s open-weight models while remaining GDPR-compliant and avoiding exposure to US data laws.
Released on August 5th, GPT-OSS are the first open-weight models from OpenAI since GPT-2, offering organisations the ability to run frontier-level AI with full control over performance, cost, and data privacy. By adding gpt-oss-120B to Hyperstack AI Studio, NexGen Cloud extends these benefits with instant, GDPR-compliant serverless inference – ideal for companies operating in stricter data security environments. Customers can choose to deploy the model directly in AI Studio for plug-and-play access or run it on dedicated Hyperstack Virtual Machines for maximum customisation and control. Hyperstack’s cost-per-minute billing keeps GPU usage transparent and predictable.
“Bringing this model to our platform within days of its release is a testament to the team’s speed and dedication to our customers. By running these models entirely on our European-owned cloud and through our own APIs – With no dependency on third-party (API) providers – we ensure organisations can maintain full control over their data, meet GDPR requirements, and avoid exposure to US data laws. With both virtual machine and serverless options, customers have plug-and-play solutions for any level of privacy, performance, and flexibility,” said Cory Hawkvelt, CPTO of NexGen Cloud.
OpenAI recently released gpt-oss-120B and gpt-oss-20B under the Apache 2.0 license. Currently, NexGen Cloud’s Hyperstack AI Studio offers the 120B model with serverless inference, giving customers instant access without infrastructure setup. Both models are accessible today via dedicated Hyperstack Virtual Machines for customers who require maximum control and customisation, with tutorials available on www.hyperstack.cloud. Fine-tuning for the 120b model will be introduced soon.
Open-weight models give organisations the power to run state-of-the-art AI without relying on proprietary APIs or SaaS-wrapped AI providers. NexGen Cloud’s Hyperstack is the first European-owned GPU cloud to enable this at enterprise scale, offering on-demand access to the latest NVIDIA GPUs, high-speed networking and ultra-fast NVMe storage. Integrating gpt-oss into Hyperstack AI Studio ensures workloads run efficiently, securely, and in full compliance with GDPR, supported by SOC 2 Type 1 certification and enterprise-grade security standards.
