ERNIE 5.0

Baidu’s unified multimodal foundation model jointly modeling text, images, audio, and video, first announced in November 2025.

LegacyLanguageReasoningMultimodalWeights not released

MODEL NUMBER283

Specifications

Only published specifications are shown.

Input modalities

Text · Image · Audio · Video

Output modalities

Text

Open weights

Status

Legacy

Baidu’s unified multimodal foundation model jointly modeling text, images, audio, and video, first announced in November 2025.

Compared with ERNIE 4.5 ↗

PreviousERNIE 4.5— · — ctx

CurrentERNIE 5.0— · — ctx

2025-11-13Announcement

ERNIE 5.0 was announced.

2026-01-15Major update

ERNIE 5.0-0110 dropped its Preview label on LMArena; this does not establish same-day availability across every channel.

Select a card to view the related model.