← Founder Notes
Archive

The $15 chip now runs a language model

Yethikrishna ROriginal on Threads

the $15 chip now runs a language model: a developer got a 28.9-million-parameter model running offline on an esp32-s3 at 9.5 tokens a second, and microsoft’s phi-4 edge hits 92% of gpt-4o’s score on a laptop chip.

the frontier stopped being about the biggest model and started being about who runs without a network.

Provenance

The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 02:16 IST.

View the original post
Embed this note
<iframe src="https://founder.myndlabs.tech/notes/embed/the-15-chip-now-runs-a-language-model-Dd9zQ13jAXW" width="480" height="420" style="border:0;max-width:100%" loading="lazy" title="The $15 chip now runs a language model"></iframe>

More notes