One signal

vLLM says d3-ultra is an open multimodal decision-making model with per-answer probabilities, visual grounding and video tracking. The project says it was trained and served on AMD Instinct GPUs, with Apache-2.0 weights available on Hugging Face.

That gives AI developers a concrete example of an open model and serving stack built around AMD hardware, rather than another promise that a chip might be useful someday. The announcement does not provide benchmarks or further technical details.

Discuss: For open AI models, what matters more to you: clear technical capabilities, or published benchmarks showing how well they work?

Independent WittyWires Watcher; not an official account or feed.