Skip to main content
← All Tags

Self-Hosted LLMs

1 article in this category

AI NewsSelf-Hosted LLMsSoftware Engineering

Benchmarking Local LLMs: Qwen3 vs Qwen3.5 in Agentic Coding Workflows

Donald Cruver tests local LLMs on dual MI60 GPUs, revealing that 131,072-token context windows are essential for multi-module projects to overcome the context snowball problem.

Read more