Models
Optimising Embedding Models for Domain-Specific Document Retrieval
Match embedding models to your document domain — code, medical, legal, or technical — for significantly better local RAG retrieval quality.
May 31, 2026 - 10 min read
01
Llama family
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
02
Mistral and Mixtral
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
03
DeepSeek reasoning
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
04
Embedding models
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
05
Vision-language models
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
06
Quantization notes
A focused resource track with implementation notes, tradeoffs, and upgrade paths.
Models
Best Embedding Models for Local RAG Systems in 2026
Compare embedding models for local retrieval-augmented generation, from BGE to E5 to Nomic Embed, and choose the right one for your document pipeline.
May 31, 2026 - 9 min read
Models
GGUF Quantisation Guide: Choosing the Right Format for Your Local LLM
Understand GGUF quantisation levels, choose Q2 through Q8 for your hardware, and balance quality against VRAM usage for every local model.
May 31, 2026 - 10 min read
Models
Gemma 3 Local Setup: Run Google's Open-Weight Models on Your Own Hardware
Install Gemma 3 locally with Ollama or Hugging Face, compare sizes, and build privacy-first workflows on Google's efficient open-weight architecture.
May 31, 2026 - 9 min read
Models
Phi-4: How Microsoft's Compact Model Changes Local AI Deployment
Run Phi-4 locally on modest hardware, understand why its small size punches above its weight, and integrate it into practical workflows.
May 31, 2026 - 8 min read
Models
Qwen 2.5 Local Setup Guide: Alibaba's Versatile Open-Weight Model Family
Install and run Qwen 2.5 models locally with Ollama or vLLM, compare size variants, and deploy them for chat, coding, and multilingual tasks.
May 31, 2026 - 9 min read
Models
DeepSeek R1 Local Setup Guide: Run a Reasoning Model on Your Own Hardware
Install DeepSeek R1 locally, configure quantised variants for consumer GPUs, and build a private reasoning workflow that keeps data off third-party servers.
May 31, 2026 - 10 min read
Models
Mistral vs Llama vs Qwen: Choosing the Best Open-Weight Model Family
Compare Mistral, Llama, and Qwen model families across performance, hardware fit, ecosystem support, and practical use cases.
May 31, 2026 - 10 min read
Models
Prompt Chaining: Connect Multiple Prompts for Better Local AI Results
Chain short, focused prompts together to produce better results than one giant instruction with local models.
May 31, 2026 - 7 min read
Models
[News] Liquid AI Releases LFM2.5-8B-A1B: On-Device MoE with 128K Context
Liquid AI's new LFM2.5-8B-A1B packs 8B total parameters (1B active) with a 128K context window, trained on 38 trillion tokens, and runs on llama.cpp, MLX, vLLM and SGLang from day one.
May 31, 2026 - 4 min read
Models
Liquid AI LFM2.5-8B-A1B: The 1B-Active-Parameter MoE Model That Runs on Consumer Hardware
Liquid AI released LFM2.5-8B-A1B, a Mixture-of-Experts model with only 1B active parameters per token, a 128K context window, and day-one support for llama.cpp — making it one of the most efficient models for local inference.
May 30, 2026 - 9 min read
Models
How to Run Liquid AI LFM2.5-8B-A1B Locally: A New MoE Model for Consumer Hardware
Liquid AI's LFM2.5-8B-A1B is an 8B-parameter MoE model with only 1B active parameters, trained on 38T tokens with 128K context — and it runs on consumer hardware via llama.cpp and GGUF.
May 30, 2026 - 10 min read
Models
[ArXiv] Qwen-VLA: Unified Vision-Language-Action Model for Robotics
Alibaba's Qwen team releases Qwen-VLA, an embodied foundation model that unifies vision, language, and continuous action generation across diverse robot platforms.
May 30, 2026 - 3 min read
Models
Build a Two-Model Workflow with a Fast Model and a Reasoning Model
Combine a small fast model and a stronger reasoning model to balance speed, cost, and quality.
May 29, 2026 - 10 min read
Models
How to Choose the Right Local Model Size
Match model size to your hardware, latency target, and task before you chase benchmark hype.
May 27, 2026 - 8 min read
Models
Quantisation Levels Explained for Real-World Local AI
Understand what 4-bit, 5-bit, and 8-bit quantisation actually mean for speed, quality, and memory.
May 26, 2026 - 9 min read
Models
Prompt Engineering for Local AI That Produces Better Answers
Use clearer instructions, better context, and repeatable prompt patterns to improve local model output.
May 24, 2026 - 8 min read
Models
Choosing the Best Embedding Model for Local Search
Compare embedding models for retrieval, semantic search, and document clustering on local hardware.
May 24, 2026 - 8 min read
Models
Best Local AI Models for Beginners
A beginner-friendly map of local model types, sizes, and practical first choices.
May 13, 2026 - 8 min read