# JiT-DDT Architecture Trains Text-to-Image Models 3.6x Faster

Digest AI · Research · published 2026-09-16T16:58:13Z

Canonical: https://digestai.news/story/jit-ddt-architecture-trains-text-to-image-models-3-6x-faster

## Summary

Researchers have introduced JiT-DDT, a novel encoder-decoder pixel-space architecture designed to accelerate text-to-image model training. Traditional Latent Diffusion Models (LDMs) rely on a separate Variational Autoencoder (VAE) and Diffusion Transformer (DiT), which limits token compression and creates training bottlenecks. By eliminating the VAE and integrating compression directly into the DiT, the new method achieves a 32x32 token reduction.

To overcome the loss of fine-grained detail common in pixel-space models, JiT-DDT splits the DiT into an encoder that predicts a low-resolution structural sketch and a decoder that reconstructs high-frequency details. When tested against the Linum v2 baseline, JiT-DDT trained a text-to-image model using 3.6× fewer GPU-hours while generating images with 4× more pixels. The team has released the code and model weights under an Apache 2.0 license.

## Key points

- JiT-DDT trains text-to-image models 3.6x faster than the Linum v2 baseline while generating four times the pixel resolution.
- The architecture eliminates the traditional VAE, pushing compression directly into the DiT to achieve 32x32 token reduction.
- The code and research model weights have been open-sourced under the Apache 2.0 license as a research artifact.

## Why it matters

This research offers a highly efficient alternative to standard Latent Diffusion Models, proving that pixel-space models can bypass VAE bottlenecks to drastically lower GPU training costs for generative image and video AI.

## Sources

1. [Training Text-to-Image Models 3.6× Faster](https://linum.ai/field-notes/jit-ddt) (linum.ai, 2026-09-16)

## Cite

Digest AI, "JiT-DDT Architecture Trains Text-to-Image Models 3.6x Faster", 16 September 2026, https://digestai.news/story/jit-ddt-architecture-trains-text-to-image-models-3-6x-faster

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/jit-ddt-architecture-trains-text-to-image-models-3-6x-faster.json