ZML Releases Free Multi-Chip AI Inference Server
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- ZML released ZML/LLMD, a free LLM inference server that runs open-source models across Nvidia, AMD, Google TPU, Apple Metal, and Intel Arc chips, with founder Steeve Morin saying the goal is to break existing silos and let enterprises hit peak performance on a mix of hardware.
- Morin said ZML has reached the point of co-designing silicon with chip partners and credited the Paris-based startup's 20-person team for shipping fast, with more releases planned.
- Morin raised $20 million from investors including Harry Stebbings' 20VC, Xavier Niel's Kima Ventures, Kindred Capital, LocalGlobe, AALVC, Drysdale Ventures, >commit, and Puzzle Ventures — leaning on his track record as VP of engineering at Zenly, which Snapchat acquired for nine figures in 2017.
- ZML/LLMD launches as a free, closed-source product — distinct from ZML's open-source inference framework released in 2024 and updated in March — so the company can study usage patterns before deciding on pricing.
- The startup enters a crowded 'inference gold rush' field that includes Baseten (recently valued at $13 billion), Inferact (from the creators of vLLM), and RadixArk (the commercial company behind SGLang).
- Morin pointed to European AI chipmakers — Axelera, Fractile, Kalray, OLIX, Q.ANT, SiPearl, SpiNNcloud, and VSORA — as potential ZML partners that could break into the market through the software.
- Notable cap-table backers include Docker/Dagger founder Solomon Hykes, Hugging Face's Clément Delangue and Julien Chaumond, and Turing Award winner Yann LeCun, now with AMI Labs.
Why it matters: Enterprises running AI workloads gain a software layer to mix Nvidia, AMD, Google, Apple, and Intel chips for inference based on cost and energy use rather than vendor lock-in — a direct pressure point on Nvidia's dominance, though Morin told TechCrunch he isn't bearish on Nvidia and has a good relationship with the chip giant. The free launch gives ZML adoption data before it commits to a paid price.
