Groq, an artificial intelligence inference platform based in the United States, has teamed up with Saudi Arabia’s HUMAIN, which is backed by the Public Investment Fund (PIF), to announce the immediate global launch of OpenAI’s newest open-source models—gpt-oss-120B and gpt-oss-20B—now available on GroqCloud. This collaboration aims to deliver real-time responses with low latency and provide local development support within the kingdom.
This rollout equips developers with extensive 128,000-token context windows, enabling real-time inferences and access to various integrated tools, including server-side code execution and web search functionalities.
The service is now operational, with local support powered by HUMAIN, recognized as Saudi Arabia’s leading provider of AI services.
Raising the Bar for Open-Source Models
According to Jonathan Ross, the CEO of Groq, “OpenAI is establishing a new high-performance benchmark for open-source models.” He continued by stating, “Groq was engineered to efficiently run such models from inception, making it accessible for developers globally. Partnering with HUMAIN enhances local access and support within Saudi Arabia.”
Accelerating AI Integration in Saudi Arabia
Tareq Amin, CEO of HUMAIN, expressed that their collaboration will significantly enhance the proliferation of AI in the region. He remarked, “Groq offers unparalleled inference speed, scalability, and cost-effectiveness necessary for introducing cutting-edge AI in the kingdom. Together, we are paving the way for a new era of innovation in Saudi Arabia.”
Both models, gpt-oss-120B and gpt-oss-20B, are engineered to operate at high throughput on GroqCloud, with the 120B model delivering over 500 tokens per second (t/s) and the 20B model exceeding 1,000 t/s.
The pricing structure is as follows:
- gpt-oss-120B: $0.15 per million input tokens, $0.75 per million output tokens
- gpt-oss-20B: $0.10 per million input tokens, $0.50 per million output tokens
For a limited time, Groq is offering a waiver on fees for tool calls utilizing OpenAI’s open models.
Groq’s global framework, which includes data centers located across North America, Europe, and the Middle East, aims to ensure high accessibility with minimal latency. Furthermore, this partnership reinforces Groq’s commitment to supporting OpenAI’s open-source endeavors, building on earlier integrations such as Whisper.
HUMAIN, which specializes in comprehensive AI solutions including data centers, large language models (LLMs), and applied AI, positions this collaboration as part of Saudi Arabia’s broader initiative to lead in artificial intelligence intellectual property and digital infrastructure.