Offline 180.9M-parameter LLM and Agent inference on ESP32-P4

developer-tools / Show HN

SCORE

A project enabling offline inference of a 180.9 million parameter large language model and agent workflows directly on the ESP32-P4 microcontroller. It runs entirely on-device without requiring a network connection or external compute. Targeted at embedded developers exploring edge AI and on-chip LLM deployment.

VISIT PRODUCT →

Sources (1)

Show HN

https://news.ycombinator.com/item?id=49183893

1 PTS

Score Breakdown

Traction

raw 1.00 · weight 35%

3.5pts

Novelty

0 days old · weight 20%

20.0pts

Source diversity

1 source · weight 10%

3.3pts

AI quality

raw 72.00 · weight 35%

25.2pts

Technically impressive niche project running 180M-param LLM on ESP32-P4; strong differentiation for edge AI devs, but zero traction signals limit confidence.

Final Score52/100