more retrocomputing experiment than useful assistant, which is exactly why it works. it knows its lane, stays small, and leans into short strange answers instead of pretending to be smarter than it is.
Z80-μLM
A 2-bit quantized conversational language model that runs on an 8-bit Z80 processor.
42 ratings4.22
Saves514
CategoryAI
PlatformWeb
PricingFree
Z80-μLM is a compact, quantized language model designed to run conversational AI on vintage 8-bit Z80 hardware with limited memory. It allows training conversational models in Python, exporting them as CP/M .COM binaries, and running them interactively on various Z80-based platforms including CP/M and ZX Spectrum. It uses trigram hash encoding and 2-bit weight quantization to produce terse, personality-driven text responses character-by-character.
Ratings & Reviews
Write a reviewInformation
CategoryAI
PlatformsWeb
PricingFree
AccountNot confirmed
ListedDecember 2025
WebVisit
Report a problem
Around the Web
Show HN514 source points · 0 comments