LLMs & Inference

4 articles

MiniCPM5-2B quantization report

MiniCPM5-2B quantization report: the best GGUF weights and K/V cache quants on llama.cpp and BeeLlama.cpp, squeezing a SOTA model into 3 GiB RAM.

Guido Imperiale

Sep 22, 2026 · 5 min read

LLMs & Inference

LLMs: Intelligence vs. cost

LLM intelligence vs cost: logarithmic scales in plots hide the cost chasm between frontier and cheap models; actual OpenRouter prices alter the Pareto frontier.

Guido Imperiale

Sep 1, 2026 · 6 min read

LLMs & Inference

Sandboxing Code Mode for Local LLM Agents

Code mode can make local LLM agents more practical, but executing model-written code brings sandboxing back into the architecture.

Nick Byrne & Khuyen Tran

May 4, 2026 · 4 min read

LLMs & Inference

What I Learned Making a Local LLM Do Real Work

Learn what makes a local LLM agent reliable enough for real work: evals, deterministic Python logic, and why a better model beats engineering.

Adam Lewis

Apr 21, 2026 · 7 min read

LLMs & Inference

Posts in LLMs & Inference

MiniCPM5-2B quantization report

LLMs: Intelligence vs. cost

Sandboxing Code Mode for Local LLM Agents

What I Learned Making a Local LLM Do Real Work