🔀

Multi-model Config

Optimised routing between multiple models for different task types.

HK$380

What is this?

The right model for every task

Not every AI task needs the same model. Quick questions need a fast, lightweight model. Complex reasoning needs a powerful one. Code generation needs a specialised coding model. Multi-model Config sets up intelligent routing so queries automatically go to the best model for the job.

You get speed when you need speed, and power when you need depth — all managed automatically through Open WebUI. No manual switching, no guesswork about which model to use.

We install and optimise multiple models on your Mac — Qwen for fast responses, Llama for deep reasoning, Codestral for code generation, and LLaVA for image understanding — then configure task-based routing so each query is handled by the most appropriate model.

The result is an AI system that feels smarter and faster than any single model alone, running entirely on your local hardware with zero cloud dependency.

How it works

Intelligent model routing

graph LR
    A[User query] --> B[Task classifier]
    B --> C{Router}
    C --> D["Fast model
Qwen"]
    C --> E["Smart model
Llama"]
    C --> F["Code model
Codestral"]
    C --> G["Vision model
LLaVA"]
    D --> H[Best response]
    E --> H
    F --> H
    G --> H

What you get

Everything included

Multiple models installed and optimised (Qwen / Llama / Codestral / LLaVA)
Task-based routing configuration
Open WebUI model switching setup
Performance benchmarks for each model
Memory and resource optimisation
Model selection guide
Quick-switch keyboard shortcuts

Who is this for?

Built for diverse workflows

⚡

Power Users

Who use AI for diverse tasks and need the right model for each one.

💻

Developers

Needing code-specialised AI alongside general-purpose models.

👥

Teams

With varied AI needs across different roles and departments.

🎯

Performance-conscious

Wanting both speed and quality without compromise.