v0.1.6 v0.1.7
low risk
weave-os/routerJan 12, 2026 → Jan 16, 20263 days between releases
What's New
Added Voyage 4 embedding models with flexible dimensions and quantization support.
New Features
- Added support for Voyage 4 embedding models (large, mid, lite, nano).
- Models now support 32,000 token context length.
- Flexible dimensions (2048/1024/512/256) via Matryoshka learning.
- Support for multiple quantization formats (float, int8, binary).
- Added specific model identifiers: voyage-4-large, voyage-4, voyage-4-lite, voyage-4-nano.
Upgrade Notes
You can now configure Voyage 4 models in your embedding settings. Ensure your environment supports the new 32k context length requirements if using these models.
The Two Releases
v0.1.6
v0.1.6
Jan 12, 2026· 9 months ago
Minimal release notes: "**Full Changelog**: https://github.com/chonkie-inc/catsu/compare/v0.1.5...v0.1.6"
Full release notesv0.1.7
v0.1.7
Jan 16, 2026· 9 months ago
This release introduces support for Voyage 4 embedding models, providing a range of sizes with flexible dimensions and context lengths.
Full release notes