DeepSeek's new bargain model accelerates AI's race to zero

Chinese AI startup DeepSeek released a powerful new coding model that costs nearly nothing — charging just twenty-eight cents for work that would run twenty-five dollars on Anthropic's Claude Opus, among the industry's most capable systems. This isn't an isolated discount; it's the latest flash in a full-scale AI price war. In July alone, OpenAI slashed its GPT Luna model by eighty percent, Google released three new flash Gemini models built for efficiency, and Meta quietly flipped to closed-source with an aggressively priced Muse Spark release. Anthropic stands nearly alone at premium pricing, betting that developers will pay extra for safety. As the gap between top-tier models shrinks, buyers are shopping on price — per Axios, a shift that could commoditize intelligence itself and threaten the pricing power that frontier labs have built their hundred-billion-dollar infrastructure bets around.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton