Project AURIONAI Use-Cases Research & Innovation
BAAI
image generation

Emu3.5

Emu3.5, a 34B native multimodal model from BAAI that predicts next tokens across vision and language, including image generation.

BAAIBAAIChina
Figure from the Emu3.5 paper: (a) Comparison with SOTA models on image generation and editing benchmarks.
Levelmodel
Classimage generation
Parameters34.1B
Contextnot in the notes
Released30 Oct 2025
Opennessunknown
Runs onnot in the notes
Statusactive

Benchmark Results

vendor-reported
GenEvalgenerative media
paper-reported, labelled unverified by Codesota; year 2025; 34B, ARvendor-reported
0.86
00.51

Score in score. Higher is better.

rank 5 of 7, Codesota aggregator30 Oct 2025 arxiv.org, codesota.com
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

R&D Topics

Unified multimodal models for understanding and generationexemplar

Identity

Developer
BAAI
Family
not in the notes
Released
30 Oct 2025 sourcesinferred from page
Weights published
–
Record checked
11 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
autoregressive Decoder-only transformer trained with next-token prediction, 64 layers; Discrete Diffusion Adaptation (DiDA) for parallel image token generation at inference
Total parameters
34.1B sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
text
Output
image
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
unknown
Licence
not in the notes sources
Access routes
not in the notes
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
not in the notes
Footprint
not in the notes
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. arxiv.org/abs/2510.26583 https://arxiv.org/abs/2510.26583

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Emu3.5", https://aurion.semai.id/models/emu3-5/, records of October 2026.