clevis@tinyweights:/home$ cat ./welcome.txt
> Small language models — Gemma, Phi, SmolLM, Qwen, and the rest — are getting surprisingly capable. Benchmarks, local deployment, quantization, and hands-on guides for running small LLMs on real hardware.
> Benchmarks, deployment guides, and hands-on for devs running AI on real hardware.
23 posts · last updated 2026-05-25 · all writing CC BY 4.0
clevis@tinyweights:/home$ ls -lh --sort=time
05-15
7min
#small-models
GGUF vs ONNX vs MLX: Which Model Format Should You Use for Local Inference?
05-14
11min
#small-models
Ollama vs LM Studio vs llama.cpp: Which Local AI Runtime Should You Use?
05-14
10min
#small-models
The Best Small Language Models in 2026: A Practical Comparison
05-13
8min
#small-models
Qwen3.5-0.8B: A Multimodal Thinking Model That Fits in 1 Gigabyte
05-11
7min
#small-models
Qwen3-Coder-Next: Run a Frontier-Level Coding Agent Locally on Consumer Hardware
04-05
6min
#small-models
Gemma 4: Taking Agentic Workflows to the Edge
03-25
8min
#small-models
Deep Dive: Running Reka Edge Locally for Frontier-Level Vision AI on Mac and PC
03-24
7min
#small-models
AI in Your Pocket: How Liquid AI’s Apollo App Lets You Run Chatbots Completely Offline
03-24
6min
#small-models
The 'Small' Model That Does It All: How Mistral Small 4's Unified Architecture Kills the Need for Specialized AI
03-23
6min
#small-models
Out of the Cloud, Into the Wild: How Small AI Models and Physical AI Are Taking Over the Edge