ollama-optimizer

Detect system hardware and generate an Ollama optimization plan.

104|15|Updated Feb 3, 2026
One-click install
npx skills add https://github.com/luongnv89/skills --skill ollama-optimizer
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ollama-optimizer
Source: https://github.com/luongnv89/skills/tree/main/skills/ollama-optimizer
Command: npx skills add https://github.com/luongnv89/skills --skill ollama-optimizer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

The Ollama optimizer skill detects your system and generates a tailored optimization plan to maximize local Ollama performance, balancing speed and memory usage.

Core Features & Use Cases

  • System detection and hardware profiling (CPU, RAM, GPU, unified memory)
  • Generated optimization plan including environment variables, model recommendations, and Modelfile tuning
  • Step-by-step execution checklist to apply optimizations across macOS, Linux, and Windows
  • Use cases: preparing a workstation for 3B-70B models, reducing memory footprint on 8-16GB machines, enabling GPU offload on Apple Silicon

Quick Start

Run the ollama-optimizer skill to detect your system and generate a tailored Ollama configuration plan.

Frequently Asked Questions about ollama-optimizer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I optimize Ollama performance for my local hardware?

To optimize Ollama performance locally, run a system detection process that profiles your CPU, RAM, and GPU, then apply generated environment variables and Modelfile tuning to balance inference speed and memory usage.

What is the best way to configure Ollama GPU offload on Apple Silicon?

Configuring Ollama GPU offload on Apple Silicon requires detecting your unified memory capacity and generating a tailored Modelfile that balances model size against available memory to maximize local inference speed.

Can I run large language models on an 8GB RAM machine using Ollama?

Running large language models on 8GB RAM machines with Ollama requires memory management tuning and specific model recommendations to reduce the memory footprint while maintaining functional local inference.

Does this Ollama optimization approach work across macOS, Linux, and Windows?

Yes, this Ollama optimization approach works across macOS, Linux, and Windows, providing a structured execution checklist with system-specific environment variable recommendations for each platform.

How do I tune an Ollama Modelfile to reduce memory footprint?

Tuning an Ollama Modelfile to reduce memory footprint involves applying hardware-specific environment variable recommendations and execution checklists generated after assessing your system dependencies and GPU capabilities.