I built a free tool to help people choose a local AI model that fits their computer

Description and use

MamiLens simplifies local AI model selection by considering user hardware and task, offering three suitable models and compatible software.

Problem

Choosing a local AI model that fits a user’s computer is difficult because model pages rarely provide direct answers about compatibility and required resources.

Outcome

MamiLens streamlines local AI model selection by providing hardware‑aware recommendations and engine guidance, cutting down setup time and reducing trial‑and‑error.

Workflow

• 1. User inputs OS, RAM, GPU and GPU memory, desired task, text volume, and speed‑vs‑capability preference.
• 2. MamiLens calculates memory estimates and selects three model options: balanced, lightweight, and high‑capability.
• 3. For each option, MamiLens shows model size, memory needs, expected fit, and evidence.
• 4. User selects a model and MamiLens recommends a compatible engine (Ollama, llama.cpp, LM Studio, etc.) based on OS, model format, and usage scenario.
• 5. MamiLens generates an installation plan, including software download links and model artifact location.
• 6. User follows the plan, runs the model, and reports success or issues to improve future recommendations.

Evidence

MamiLens connects the confirmed model to the software step. It considers the operating system, model format, technical comfort, intended use and number of people who may use the system at the same time.

Source