---
format: "aidr-story-markdown/v1"
id: "aa65eb3ab36b5eef10afd8013f57ac938e7f406160eadcec57c2cf92b2630f43"
canonical_url: "https://aidr.today/aa65eb3a?lang=en"
title: "The Design Space of Tri-Modal Masked Diffusion Models"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-14T00:00:00.000Z"
category: "Research"
topics: ["llm"]
source_urls: ["https://machinelearning.apple.com/research/tri-modal-masked-diffusion"]
summary: "Researchers introduced the first tri-modal Masked Diffusion Models (MDM) pretrained from scratch on text, image-text, and audio-text data, analyzing multimodal scaling laws and achieving strong results at a 3B model scale."
---

# The Design Space of Tri\-Modal Masked Diffusion Models

> [Open the canonical story](<https://aidr.today/aa65eb3a?lang=en>)

**Published:** 2026-10-14T00:00:00.000Z
**Category:** Research
**Topics:** llm

## Summary

Researchers introduced the first tri\-modal Masked Diffusion Models \(MDM\) pretrained from scratch on text, image\-text, and audio\-text data, analyzing multimodal scaling laws and achieving strong results at a 3B model scale\.

## Sources

- [Story source](<https://machinelearning.apple.com/research/tri-modal-masked-diffusion>)