France’s ZML Wants to Break Nvidia Lock-in with Free Cross-Chip AI Software
A Paris startup wants to loosen Nvidia’s grip on AI, not with a new chip, but with software. ZML has released a free tool that runs open-source models fast across Nvidia, AMD, Google, Apple, and Intel silicon alike.
July 8, 2026 – 4:37 pm
Nvidia still rules AI hardware, but its walls are thinning. ZML, a Paris startup backed by AI pioneer Yann LeCun, has released free software that runs open-source language models across a mix of chips, as reported by TechCrunch. The list includes Nvidia, AMD, Google’s TPUs, Intel, and Apple.
The tool, ZML/LLMD, is an inference server. Inference means running a trained model to answer prompts, the part of AI that now consumes most compute. Founder Steeve Morin explains that the goal is to break the silos that lock users to one vendor and squeeze each chip to its top speed.
Why a mix of chips matters
Cost is the primary driver. As AI bills climb, enterprises and clouds seek freedom to pick cheaper or less power-hungry silicon for specific tasks. “The idea is to give people back the power to create their own system,” Morin said. Doing so effectively could also spark a wave of novel chipmakers, many based in Europe. Morin mentions Axelera, Fractile, Kalray, SiPearl, VSORA, and others. Software that treats their chips as first-class, not second-best, gives buyers a compelling reason to try them.
A crowded, costly race
Morin acknowledges Nvidia’s position but emphasizes ZML’s positive relationship with the chip giant. However, the competition is fierce. The “inference gold rush” has birthed rivals like Baseten, valued at $13bn, plus teams behind open-source projects vLLM and SGLang. All chase the same objective: making AI cheaper to run.
Morin believes ZML stands out. “We have reached the point where we are co-designing silicon,” he said. His lean team of 20 has moved swiftly, with more releases on the horizon.
Why it matters
LLMD is currently free to encourage usage, not yet a paid product. Its significance lies in its origin. A tool designed to loosen Nvidia’s grip and back Europe’s AI stack emerged from Paris, not Silicon Valley. Morin, who secured $20m from investors including Xavier Niel’s Kima Ventures, expresses his location clearly: “I couldn’t have done ZML anywhere but in Paris.”