Model family

Q: How deeply does Build Applications with Local AI Models on a Mac cover model family?

Compares family origins, strengths, sizes, and uses but does not explain model internals.

As explained in Build Applications with Local AI Models on a Mac

A group of related language models that share a common architecture, training lineage, or design.

A model family is a product line of related language models from a company. Families differ in design philosophy, training data, intended uses, and strengths, helping you compare options for local use. section 12.1

Table 4.2 Major open source model families available through Ollama

Model family	Creator	Notable sizes	Strengths	Best for
Gemma 3	Google DeepMind	4B, 12B, 27B	Efficient, strong reasoning	General tasks, learning
Llama 4 [3]	Meta	Scout, Maverick	Versatile, large community	General-purpose tasks, coding
Qwen 3.5 [4]	Alibaba Cloud	4B, 8B, 14B, 32B, 72B	Multilingual, strong at math	Multilingual tasks, math
Mistral [5]	Mistral AI	7B, 8x7B (Mixtral)	Fast inference, efficient	Speed-critical applications
Phi-3	Microsoft	3.8B, 14B	Compact yet powerful	Lightweight deployments

Compare creators, sizes, strengths, and suggested uses across five open-source families available through Ollama. section 4.4.1

A starting comparison

For learning, start with Gemma 3 (4B), then try Qwen 3.5 (4B) and compare their responses. section 4.4.1

How Build Applications with Local AI Models on a Mac covers it

Cover of Build Applications with Local AI Models on a Mac

liveBook manning.com

Read Build Applications with Local AI Models on a Mac on liveBook Get the book at manning.com

Related concepts

Gemmaincludes · Gemma is a family of open-weight generative AI models developed by Google.
Mistralincludes · Mistral is a family of large language models developed by Mistral AI.
Qwenincludes · Qwen is a family of pretrained generative language models developed by Alibaba Cloud, with variants for different tasks and modalities.
Llamaincludes · Llama is a family of autoregressive large language models developed by Meta and released in multiple versions and sizes.
parameter counthas part · The number of adjustable numerical parameters in a model, typically learned or tuned during training.

This explainer was generated by Manning's knowledge graph from the text of Build Applications with Local AI Models on a Mac. The section links point to the parts of the book it draws on.