What Humain just rolled out
Saudi Arabia's Humain, backed by the Public Investment Fund, says it has built a national AI platform using an open model from China's MiniMax Group Inc. On Thursday, it also introduced humain-m3, an Arabic-language model derived from MiniMax's M3. The company says the goal is to lift up Arabic AI and add to the broader progress of frontier intelligence.
Performance, access and how to use it
According to Humain, Humain-M3 posted the top average performance when compared with frontier systems on seven public Arabic benchmarks. Developers and companies can get to the model via Humain Node, which serves as a portal to advanced models. CEO Tareq Amin said the firm partnered tightly with MiniMax to commission and build the system, and noted the company has also worked with Mistral and Cohere.
"Arabic is spoken by hundreds of millions of people, yet it remains significantly underrepresented at the frontier of artificial intelligence," Amin said in a statement. "With humain-m3, we are investing in changing that." He added on LinkedIn: "This is not about importing a model and putting a Humain label on it."
When technology changes fast, steady investing still matters, so get our free Always Be Buying E-Book
The bigger sovereignty push
In January, South Korea's government dropped Naver Corp. from a shortlist for creating a national AI model, stating that it discovered a Chinese vision encoder had been used in part in its contest submission. Saudi Arabia joins countries from France to India intent on securing a place in a fast-moving field that US and Chinese firms increasingly dominate, at a time when leaders such as Anthropic are curtailing or restricting access to their models. Humain has poured money into computing facilities and semiconductors, aiming to make the kingdom a contender on the world stage in the AI race.
For anyone building products in Arabic, the headline is simple: a new Arabic-focused model is live and reachable through Humain Node. That can expand the menu of tools available to local teams and businesses looking to put AI to work in the languages their customers actually use.
