Starstruk Logo MarkStarstruk

v0.32.1

wxt-dev/wxtv0.32.1Jul 16, 2026by github-actions[bot]

AI Summary

This release focuses on improving reasoning capabilities for Gemma 4, fixing memory leaks in the MLX cache, and enhancing the interactive agent's context.

Key Highlights

  • Improved Gemma 4 tool calling and multi-turn reasoning reliability
  • Fixed a recurrent MLX model cache leak that could increase memory use
  • Enhanced interactive agent context with current working directory
  • Fixed `ollama launch` model picker for deprecated models

Full Release Notes

## What's Changed
- Improved Gemma 4 tool calling and multi-turn reasoning, including more reliable tool-response continuations
- Fixed a recurrent MLX model cache leak that could increase memory use across requests, and improved cache snapshot performance
- MLX text model loading now respects `OLLAMA_LOAD_TIMEOUT`
- Agent web search and fetch now tell users to run `ollama signin` when authentication is required
- The interactive agent now receives the current working directory for better project context
- Fixed `ollama launch` so choosing **Pick another model** for a deprecated model passed with `--model` opens the model picker
- Updated VS Code setup documentation for the official Ollama extension

**Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.0...v0.32.1-rc0