Gemma 4 MLX Challenge Boosts Local Inference to 178 tok/s
TheMoonMidas · x · 2026-08-28
The Gemma 4 MLX Challenge has achieved a new speed record on Apple Silicon, pushing inference speeds from a baseline of 153.1 tok/s to 178 tok/s. Driven by @yukonresearch and @eigenlabs, the competition focuses on optimizing the MLX framework for local AI efficiency. The current top submission achieved a 16.6% overall improvement.
Related event: Gemma 4 MLX Challenge Boosts Mac Inference to 178 tok/s(2 posts)→
More from Infra
- Magnitude open-source inference engine beats Ollama by over 2x with auto-tuned kernels — nickbaumann_ · 2026-10-15
- Nvidia to announce $28B in AI investments in Europe this week, report says — georgemillo · 2026-10-15
- A $10,811 Cloudflare Bill From an Unused AI-Written Test Project: Refunds Aren't a Safety Net — sujingshen · 2026-10-15
- Retraining-Free Method Turns Transformers Into Elastic Models Adapting to Any Compute Budget — VISLab-Amsterdam · 2026-10-15
- vLLM's MTP3 validation: AIME25 up to 0.9667, prefill kernels up to 84% lower latency — vllm_project · 2026-10-15
- Open-weights s1-llm-auto-router picks the right LLM in 26ms on a CPU — airesearch12 · 2026-10-15