Ollama supports the new checkpoint format
The story so far
Ollama has released version 0.40.9-rc3, replacing the pre-release checkpoint path with Gradium's formal (GA) format, including default voice import and voice tuning. This change enables Speech Ollama to use Phonon to generate voice on Apple Silicon and stream audio via the OpenAI-compatible /v1/audio/speech endpoint. Additionally, the d3 decision model supports parameter sizes from 2B to 27B for classification, scoring tasks, etc. Meanwhile, Flint has been extended to Linux and Windows platforms, and the failure reporting mechanism for streaming OpenAI and Anthropic messages has been improved. The ollama launch command passes the context length and capabilities of cloud models to coding tools, optimizing the thinking settings for Anthropic-compatible searches to avoid failures.
AI-written from the reports · Updated 25 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- Ollama releasesv0.40.9-rc3: Support GA Gradium Phonon checkpoints (#19227)
Speech Ollama can use Phonon to generate speech on Apple Silicon, streaming audio via the /v1/audio/speech endpoint; the d3 decision model supports parameter sizes from 2B to 27B for classification, scoring, and other tasks; Flint expanded to Linux and Windows platforms, other improvements include error reporting and model capability transfer.
- Ollama releasesv0.40.9
Ollama v0.40.9 released, running on Linux and Windows platforms, improves streaming processing failure reports for OpenAI and Anthropic messages, passes cloud model context length and capabilities to coding tools via ollama launch, optimizes thinking settings for Anthropic-compatible search to avoid failures.