• your topics
• add yours

The State of LLMs 2025: Progress, Progress, and Predictions

LLM Efficiency: From Hyperscale Optimizations to Universal Deployability

Show HN: Bushchat – open-source graph LLM interface

Show HN: Terminalot – A local-first, open-core SSH terminal with AI copilot

Show HN: I built a CLI to dump all JavaScript/CSS/assets from any webpage

Show HN: I built a standalone WASM playbook builder for sports coaches

Show HN: LLMRouter – Stop using GPT-4/o1 for everything (16 routing strategies)

Show HN: Reko – Local-First YouTube-to-Markdown LLM Summarizer

Have LLMs improved for Swift coding in the last 12 months?

Threads – A context strategy for humans and LLMs

LLMs have bad taste at where to draw abstraction boundaries

Show HN: CATArena – Evaluating LLM agents via dynamic enviroment interactions

Reflections on a Year of Prolog and LLMs

Why Enterprises Cannot Disclaim Consumer Harm Caused by LLM "Optimization"