Faster Gemma 4 on MLX with multi-token prediction
Ollama ·
Summary: Gemma 4 is now significantly faster in Ollama 0.31 on Apple Silicon via multi-token prediction (MTP), powered by MLX. Performance is now up to 90% faster when used with coding agents, as measured using the Aider polyglot benchmark.
Read the full story on OllamaExplore more: Today in AI · Research papers · AI tools · Learn AI
Related stories
- AI: Reset to Zero · 2 Nov 2026‘Coming for the King’ in AI. OpenAI, DeepSeek & Samsung. ARD #197
- TahawulTech.com · 2 Nov 2026Mistral AI enters cybersecurity deal with Orange
- Unite.AI · 2 Nov 2026Envision Deploys SST-Based 800V DC Power at Ulanqab AI Data Center
Envision said on Nov. 2, 2026, that an integrated 800-volt direct-current (800V DC) power system is now powering live computing workloads at its Galaxy AI data center campus in Ulanqab, China, a deployment the company described in its announcement as the world s first commercial
- Notebookcheck · 2 Nov 2026Meta puts its AI on two legs – with help from China
- BigGo Finance · 2 Nov 2026NVIDIA Releases 9.61 Million-Parameter Noise Removal AI for Free, Clears Multilingual Audio
- Drug Target Review · 2 Nov 2026How AI is changing the way therapeutic antibodies are designed
Who reported “Faster Gemma 4 on MLX with multi-token prediction”?
Ollama, on 29 Jun 2026. The full article is at the source link on this page.