ERNIE-ViLG

Closed weights Baidu 10B parameters December 2021

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Baidu
Organisation type
Industry
Country
China
Published
31 December 2021
Authors
Han Zhang, Weichong Yin, Yewei Fang, Lanxin Li, Boqiang Duan, Zhihua Wu, Yu Sun, Hao Tian, Hua Wu, Haifeng Wang

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Multimodal, Image generation, Vision, Language
Task
Vision-language generation, Image generation, Text-to-image, Image captioning, Language modeling/generation, Visual question answering
Approach
Self-supervised learning
Numerical format
FP16

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
10B

"To explore the landscape of large-scale pre-training for bidirectional text-image generation, we pre-train a 10-billion parameter model on a large-scale dataset of 145 million high-quality Chinese image-text pairs."

Training data
tokens

To explore the landscape of large-scale pre-training for bidirectional text-image generation, we pre-train a 10-billion parameter model on a large-scale dataset of 145 million high-quality Chinese image-text pairs.

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased
Training code
Unreleased

it seems that the model was completely substituted by its successor ViLG 2.0 - I haven't found niether model weights nor code or API.

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Foundation model
Yes
Why it is tracked
SOTA improvement

"we train a 10-billion parameter ERNIE-ViLG model on a large-scale dataset of 145 million (Chinese) image-text pairs which achieves state-of-the-art performance for both text-to-image and image-to-text tasks" "chieves state-of-the-art performance for both text-to-image and image-to-text tasks, obtaining an FID of 7.9 on MS-COCO for text-to-image synthesis and best results on COCO-CN and AIC-ICC for image captioning"

Citations
70

Sources

Where this record came from and when it was last checked.

Reference
ERNIE-ViLG: Unified Generative Pre-training for Bidirectional Vision-Language Generation
Last updated
25 May 2026

What the numbers mean

Background

ERNIE-ViLG was published by Baidu, in the country recorded as China, during December 2021. The publishing organisation is categorised as industry.

It works in the domain of Multimodal, Image generation, Vision, Language, and is recorded as performing the task of vision-language generation, Image generation, Text-to-image, Image captioning, Language modeling/generation, Visual question answering.

Its weights were never published, so it can only be reached through its provider. No graphics card changes that.

Training and provenance

The reason it appears in this catalogue at all: sOTA improvement.

Answers

ERNIE-ViLG — common questions

01

ERNIE-ViLG— who created it?

It was published by Baidu, based in China, an organisation categorised as industry.

02

ERNIE-ViLG— when was it released?

It was published in December 2021. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

03

ERNIE-ViLG— what is it used for?

It works in the domain of Multimodal, Image generation, Vision, Language, and is recorded as handling the task of vision-language generation, Image generation, Text-to-image, Image captioning, Language modeling/generation, Visual question answering. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

04

ERNIE-ViLG— what GPU do I need to run it?

None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

05

ERNIE-ViLG— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

06

ERNIE-ViLG— how many parameters does it have?

It has a parameter count of 10B. "To explore the landscape of large-scale pre-training for bidirectional text-image generation, we pre-train a 10-billion parameter model on a large-scale dataset of 145 million high-quality Chinese image-text pairs.". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

Source

Original publication

Record last updated 25 May 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.