The new open-source models from OpenAI, gpt-oss-120B and gpt-oss-20B, have been fully implemented by Saudi Arabia's HUMAIN on Groq's ultra-high-speed inference platform.
The models are housed in HUMAIN's independent AI data centers located within the Kingdom, guaranteeing complete adherence to regional legal and data sovereignty standards.
With the deployment, Saudi developers, businesses, and government agencies can access OpenAI's most cutting-edge open-source capabilities, which provide fast, low-latency inference while adhering to the Kingdom's privacy and legal requirements.
Unprecedented scale, 128K context windows, and integrated tools for semantic search and real-time code execution are features of the gpt-oss-120B and gpt-oss-20B models.
Operating on Groq's infrastructure at speeds of more than 500 and 1,000 tokens per second, respectively, they allow for sophisticated reasoning and conversation at previously unheard-of speeds.
First announced in May 2025, the news represents a new stage in HUMAIN and Groq's strategic partnership.
It establishes Saudi Arabia as a key worldwide hub for AI innovation, connecting capability, compute, and compliance throughout the GCC, the Levant, Africa, Asia, and beyond.
