Open sourceOpen-source releaseMLOps, Tooling & Infrastructure1 source · Sep 2, 2026

ollama/ollama v0.33.3

gemma4 now supports images and audio on MLX engine

llama.cppOllamaMLX
ProofCode ↗
Read at GitHub: ollama/ollama ↗← Today

Key points

  • Honor GGUF model defined default parameters

Sources (1)

  • [1]ollama/ollama v0.33.3
    GitHub: ollama/ollama · Sep 2, 12:11 AM
    * gemma4 now supports images and audio on MLX engine
    * Honor GGUF model defined default parameters

Extractive summary: sentences quoted from the sources.

Before this

  1. Aug 26, 2026ollama/ollama v0.33.1
  2. Aug 21, 2026ollama/ollama v0.33.0
  3. Aug 19, 2026ollama/ollama v0.32.15
  4. Aug 14, 2026ollama/ollama v0.32.12
  5. Aug 14, 2026ollama/ollama v0.32.11
  6. Jun 9, 2026Introducing Gemma 4 12B: a unified, encoder-free multimodal model

Related

GitHub26d ago

ollama/ollama v0.34.4

GitHub24d ago

ollama/ollama v0.40.0

GitHub5d ago

ollama/ollama v0.40.5-rc0: v0.40.5

GitHub11d ago

ollama/ollama v0.40.1