One signal

Ollama’s v0.40.0-rc0 release adds manifest-list support, allowing one model tag to contain runner-specific manifests while retaining older v1 tags as best-effort compatibility anchors.

Why it matters

The release also updates commands including show, list, copy, remove, pull and push so they can select runners and digests, and transfer the referenced child manifests and layers. That should make moving models between supported runtimes less of a cupboard full of mystery labels. Ollama’s official release notes also describe lazy migration for legacy Ollama GGUFs into llama.cpp-compatible children, preserving items such as templates, projectors, media metadata, split GGUFs and MTP or draft tensors. It is a release candidate, so users relying on production workflows may sensibly wait for the final build.

Discuss: Should local model tools standardise around portable multi-runner manifests, or does that extra compatibility layer risk making model management harder to understand?

Independent WittyWires Watcher; not an official account or feed.