Ollama announced version 0.40.7‑rc0 on 18 Oct 2026. The update lets non‑thinking models send OpenAI‑style responses, lets the llama‑server infer speculative draft types, and adds support for the PPLX Decider v1.2. The changes are listed in the GitHub release notes.
Why it matters
Ollama users can now integrate OpenAI‑compatible output without switching tools, simplifying mixed‑model workflows.