Imagenhub: Benchmark conditional image-generation models

Understand the project

Practical overview

Use Imagenhub when you need a shared process for generating and evaluating outputs across conditional image-generation models.

Researchers comparing conditional image-generation models.
Developers building repeatable image-generation or editing evaluation workflows.

Before you choose

The repository provides installation through PyPI or from source and identifies the project as MIT-licensed.

Model integrations can have different setup requirements.
The README specifies Hugging Face CLI login for some Stable Diffusion-family models and separate setup guidance for DALL-E, DreamEdit and BLIPDiffusion.
The documented benchmark example uses a text-to-image configuration; other workflows may require their corresponding configuration and data.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Generate benchmark outputs with the documented text-guided configuration and create an HTML visualization index.

Imagenhub installed from PyPI or source.
For some models, model-specific setup is required; Stable Diffusion, SDXL and CosXL require Hugging Face CLI login.

Run the command below.

Documented command

python3 benchmarking.py -cfg benchmark_cfg/ih_t2i.yml

Run the command below to produce an HTML index.
Documented command
```
python3 visualize.py --cfg benchmark_cfg/ih_t2i.yml
```

Benchmark output files organized by experiment and model, plus an HTML visualization index.

The README notes that some models require extra setup; it does not provide a single setup procedure for every integration.

Evidence and transparency

Source record details

Source availability: Active
Sources checked: 17 October 2026
Original website title: TIGER-AI-Lab/ImagenHub · Official homepage title · retained in its source language