LLM News for 2026-04-06

← 2026-04-05 2026-04-07 →

Caveman: Why use many token when few token do trick

688 points | by tosh | comments

Eight years of wanting, three months of building with AI

592 points | by brilee | comments

Codex pricing to align with API token usage, instead of per-message

195 points | by ccmcarey | comments

Running Gemma 4 locally with LM Studio's new headless CLI and Claude Code

177 points | by vbtechguy | comments

Nanocode: The best Claude Code that $200 can buy in pure JAX on TPUs

154 points | by desideratum | comments

Qwen-3.6-Plus is the first model to break 1T tokens processed in a day

43 points | by Alifatisk | comments

OpenAI's fall from grace as investors race to Anthropic

26 points | by 1vuio0pswjnm7 | comments

Show HN: TermHub – Open-source terminal control gateway built for AI Agents

5 points | by duo121 | comments