# new research: A tiny 109M-parameter model outperforms models up to 72B parameters at image-text alignment, using 42× less training data — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/8fc37a5586cbcea8

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-11-01 10:24 ingested 2026-11-02 02:41

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xkzony/new_research_a_tiny_109mparameter_model/)

## retained text

No document body was retained for this record — only its metadata was stored.

document fetched 2026-11-02 00:36 · canonical https://reddit.com/r/LocalLLaMA/comments/1xkzony/new_research_a_tiny_109mparameter_model
