# mtmd: enable video input support for LFM3 model family by tdakhran · Pull Request #31283 · ggml-org/llama.cpp — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/b7dcf69d69b5b5e5

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-28 12:16 ingested 2026-10-29 02:42

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xhmgql/mtmd_enable_video_input_support_for_lfm3_model/)

## retained text



```
now you can show your videos to tiny LFM3 models
```



document fetched 2026-10-29 00:44 · canonical https://reddit.com/r/LocalLLaMA/comments/1xhmgql/mtmd_enable_video_input_support_for_lfm3_model
