French AI Startup ZML Launches Free Inference Server to Run LLMs Across Multiple Chip Brands
Paris AI startup ZML launched ZML/LLMD in July 2026, a free inference server for open-source LLMs that runs on Nvidia, AMD, Google TPU, Apple Metal, and Intel Arc to reduce vendor lock-in. It’s not open source, following a 2024 ML inference framework updated in March, and ZML will only consider paid tiers after analyzing usage.