# ERNIE-Image — Baidu's Open-Weight Text-to-Image Diffusion Transformer

> Baidu's open-weight 8B diffusion transformer for text-to-image, with a Turbo variant that generates in 8 steps.

- Org: Baidu
- Date: 2026-04-15
- Category: model, tool
- Source: https://github.com/baidu/ERNIE-Image
- Canonical page: https://ai-tldr.dev/releases/baidu-ernie-image/

## In short

Baidu open-sources an 8B Diffusion Transformer for text-to-image under Apache 2.0, with a Prompt Enhancer for complex layouts. A fast Turbo variant (8 inference steps) uses DMD and RL distillation.

## What it is

ERNIE-Image is Baidu's first standalone open-weight text-to-image model. It uses a single-stream Diffusion Transformer (DiT) with 8B parameters and focuses on two known hard problems: accurately rendering dense text inside images (posters, infographics) and following complex multi-object instructions. Released under Apache 2.0 with ComfyUI, Diffusers, and SGLang support.

## How it works

The DiT uses Flux's VAE for image encoding and Ministral 3.3B as the text encoder. A lightweight Prompt Enhancer expands brief inputs into richer structured descriptions before the DiT runs. A companion Turbo variant distilled via DMD and RL reduces generation to 8 inference steps while preserving quality. Requires a GPU with 24GB VRAM.

## Why it matters

Text rendering inside generated images is a persistent weak spot for open-weight models. ERNIE-Image scores 0.9733 on LongTextBench and 0.8856 on GENEval, competitive with models several times its size. Apache 2.0 and native ComfyUI/Diffusers support mean it drops directly into existing pipelines.

## Who it's for

Developers building image generation pipelines that need accurate in-image text or complex layout control.

## Try it

from diffusers import ErnieImagePipeline

## Links

- [Repo](https://github.com/baidu/ERNIE-Image)
- [Model Card](https://huggingface.co/baidu/ERNIE-Image)
- [Turbo Variant](https://huggingface.co/baidu/ERNIE-Image-Turbo)