Ollama releases: v0.33.1

TL;DR

Ollama released version 0.33.1, which includes updates to Qwen3.8 Flash Next support, cmake patches, and mlxrunner structured output for improved model loading times, highlighting ongoing development and community contributions.

Detailed Summary

Ollama released version 0.33.1, updating MLX to support Qwen3.8 Flash Next, enhancing cmake compatibility, and improving mlxrunner with structured output and timeout avoidance features. This update also welcomed a new contributor. The broader impact includes improved model loading speed and reliability for users of the Ollama platform.

Key Points

• • MLX supports Qwen3.8 Flash Next
• • cmake updates for idempotent external compat patches
• • Structured output support added to mlxrunner
• • Metal GPU timeouts reduced when loading models from slow storage