If keeping up with AI felt harder than usual in early September, that is because it was. In one week, the four biggest AI labs all released new models, CNBC reports.
- Tuesday: Anthropic released Claude Fable 5.1 and Claude Mythos 5.1.
- Wednesday: Meta announced Muse Spark 1.3 and Google unveiled Gemini 3.8 Flash.
- Thursday: OpenAI released GPT-6 Astra, and Abu Dhabi's MBZUAI released its open-source K2 Horizon models.
Nvidia also agreed to buy the open-source AI platform Hugging Face for $12.9 billion, CNBC reports.
Too many models to test
"I feel like model fatigue is a real thing," Zhen Lu, CEO of AI startup Runpod, told CNBC. "We are in an environment where there's just so much frothiness that you have to make noise."
For the companies paying for AI, each launch means another round of comparing cost and quality. Suresh Vasudevan, CEO of Clockwork Systems, said that if his team wants to evaluate ten models for a task, it may only test five. "It's really challenging to go evaluate every one of the ones that are coming out right now," he said.
Not every launch is a leap
Noah Faro, technology chief at AI finance startup Farsight, told CNBC that the Anthropic, Meta and Google releases were point releases, upgrades to existing models, unlike OpenAI's Astra.
OpenAI CEO Sam Altman told CNBC that "we're all moving to faster cadences." Gartner expects $2.59 trillion in AI spending this year, a 47% rise on 2025, which helps explain the rush.
