ERNIE Image
Baidu · baidu/ERNIE-Image
Weights license: Apache 2.0 · code and components may use other terms
Quality
No comparable score
Artificial Analysis score, rank and sample count excluded: its free tier is internal use only. Source checked 2026-09-22. Leaderboard (link only)
Reported by the publisher
- Single Object
- 1.000ERNIE-Image (w/o PE) · GENEval
- Two Object
- 0.960ERNIE-Image (w/o PE) · GENEval
- Counting
- 0.778ERNIE-Image (w/o PE) · GENEval
- Colors
- 0.928ERNIE-Image (w/o PE) · GENEval
- Position
- 0.855ERNIE-Image (w/o PE) · GENEval
- Attribute Binding
- 0.793ERNIE-Image (w/o PE) · GENEval
- Overall
- 0.886ERNIE-Image (w/o PE) · GENEval
- Single Object
- 0.991ERNIE-Image (w/ PE) · GENEval
Every reported result
- Two Object
- 0.960ERNIE-Image (w/ PE) · GENEval
- Counting
- 0.819ERNIE-Image (w/ PE) · GENEval
- Colors
- 0.883ERNIE-Image (w/ PE) · GENEval
- Position
- 0.863ERNIE-Image (w/ PE) · GENEval
- Attribute Binding
- 0.723ERNIE-Image (w/ PE) · GENEval
- Overall
- 0.873ERNIE-Image (w/ PE) · GENEval
- Alignment
- 0.868ERNIE-Image (w/ PE) · OneIG-EN
- Text
- 0.979ERNIE-Image (w/ PE) · OneIG-EN
- Reasoning
- 0.357ERNIE-Image (w/ PE) · OneIG-EN
- Style
- 0.431ERNIE-Image (w/ PE) · OneIG-EN
- Diversity
- 0.241ERNIE-Image (w/ PE) · OneIG-EN
- Overall
- 0.575ERNIE-Image (w/ PE) · OneIG-EN
- Alignment
- 0.891ERNIE-Image (w/o PE) · OneIG-EN
- Text
- 0.967ERNIE-Image (w/o PE) · OneIG-EN
- Reasoning
- 0.295ERNIE-Image (w/o PE) · OneIG-EN
- Style
- 0.447ERNIE-Image (w/o PE) · OneIG-EN
- Diversity
- 0.169ERNIE-Image (w/o PE) · OneIG-EN
- Overall
- 0.554ERNIE-Image (w/o PE) · OneIG-EN
- Alignment
- 0.830ERNIE-Image (w/ PE) · OneIG-ZH
- Text
- 0.954ERNIE-Image (w/ PE) · OneIG-ZH
- Reasoning
- 0.306ERNIE-Image (w/ PE) · OneIG-ZH
- Style
- 0.434ERNIE-Image (w/ PE) · OneIG-ZH
- Diversity
- 0.248ERNIE-Image (w/ PE) · OneIG-ZH
- Overall
- 0.554ERNIE-Image (w/ PE) · OneIG-ZH
- Alignment
- 0.842ERNIE-Image (w/o PE) · OneIG-ZH
- Text
- 0.898ERNIE-Image (w/o PE) · OneIG-ZH
- Reasoning
- 0.266ERNIE-Image (w/o PE) · OneIG-ZH
- Style
- 0.421ERNIE-Image (w/o PE) · OneIG-ZH
- Diversity
- 0.177ERNIE-Image (w/o PE) · OneIG-ZH
- Overall
- 0.521ERNIE-Image (w/o PE) · OneIG-ZH
- LongText-Bench-EN
- 0.980ERNIE-Image (w/ PE) · LongTextBench
- LongText-Bench-ZH
- 0.966ERNIE-Image (w/ PE) · LongTextBench
- Avg
- 0.973ERNIE-Image (w/ PE) · LongTextBench
- LongText-Bench-EN
- 0.968ERNIE-Image (w/o PE) · LongTextBench
- LongText-Bench-ZH
- 0.959ERNIE-Image (w/o PE) · LongTextBench
- Avg
- 0.964ERNIE-Image (w/o PE) · LongTextBench
Values as printed on the card, with the publisher's own settings (card section: Benchmark › GENEval; Benchmark › OneIG-EN). Not comparable across publishers and not used in any ranking. Model card
Runs on
Enough memory 22
A publisher documents a path within this memory, on this platform (CUDA or Apple MPS). Conditional paths need offload or quantization.
Show all 22
One memory tier short 16
The same machines with the next memory size up would have enough.
Show all 16
Hardware profiles grouped by the publisher's documented minimum memory path and platform. Not a measured run.
Consider ERNIE Image from Baidu for the text-to-image task (8B parameters).
The catalog lists the publisher's documented memory path under Runs on and has not reviewed other configurations. Actual memory depends on precision, resolution, duration and offload; consult the official source.
Measured speed
No published measurement for this checkpoint
Generation time per image is shown only when a source publishes it with hardware, build and setup. It is never estimated.
How to run
Publisher-documented memory path: Official consumer-GPU requirement. Actual memory depends on precision, resolution, duration and offload; consult the official source. Official hardware evidence
Documented ways to run it
Copied as printed in the linked source (read 2026-10-18); not run by this site.
import torch
from diffusers import ErnieImagePipeline
pipe = ErnieImagePipeline.from_pretrained(
"Baidu/ERNIE-Image",
torch_dtype=torch.bfloat16,
).to("cuda")
image = pipe(
prompt="This is a photograph depicting an urban street scene. Shot at eye level, it shows a covered pedestrian or commercial street. Slightly below the center of the frame, a cyclist rides away from the camera toward the background, appearing as a dark silhouette against backlighting with indistinct details. The ground is paved with regular square tiles, bisected by a prominent tactile paving strip running through the scene, whose raised textures are clearly visible under the light. Light streams in diagonally from the right side of the frame, creating a strong backlight effect with a distinct Tyndall effect—visible light beams illuminating dust or vapor in the air and casting long shadows across the street. Several pedestrians appear on the left side and in the distance, some with their backs to the camera and others walking sideways, all rendered as silhouettes or semi-silhouettes. The overall color palette is warm, dominated by golden yellows and dark browns, evoking the atmosphere of dusk or early morning.",
height=1264,
width=848,
num_inference_steps=50,
guidance_scale=4.0,
use_pe=True # use prompt enhancer
).images[0]
image.save("output.png")git clone https://github.com/sgl-project/sglang.gitsglang serve --model-path baidu/ERNIE-ImageEvidence level: Publisher documentation linked; no execution reproduced by this site.
Before downloading
- Check the exact checkpoint and license linked here.
- Confirm a compatible runtime and precision or quantization for your platform.
- Allow for context, batch size and any extra media components beyond the weights.
Technical details
What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.
- Checkpoint / version
- baidu/ERNIE-Image
- Task and modality
- text-to-image · image
- Count and basis
- 8B · manufacturer model card
- Released
- baidu/ERNIE-Image commit history (GitHub), commit d39288b "init ERNIE-Image Resource" (README, Released Versions) · read 2026-10-19source
- Published repository size
- 31.6 GBNot runtime memory or a minimum download.files
- Published components
- Technical details pending review
- Runtime and precision
- No local execution verified.
- Commercial use
- Allowed Apache License 2.0 text (SPDX) · read 2026-10-18license text
License, parameter count and release date come from the publisher's sources listed on this page (a date no source states is left empty). The hardware requirement has not been reviewed, so memory is a labelled estimate. The repository size, when shown, is the full published download.
Sources
Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.
- Publisher checkpoint and model card
- Weight license and use terms
- Quality benchmark source · Artificial Analysis image · excluded (free tier is internal use only)
- Publisher hardware documentation
Catalog reviewed: 2026-09-22.






























