# Qwen publishes Qwen-Image-Bench evaluation model

> Qwen released a model repository for Qwen-Image-Bench, a judge/evaluation model for image-text-to-text tasks, fine-tuned from Qwen3.6-27B under Apache 2.0. The source does not assert capability, benchmark results, or licensing beyond the HF tags.

| | |
|---|---|
| **Tool** | Qwen |
| **Version** | — |
| **Kind** | release |
| **Published** | 2026-05-21 |
| **Observed** | 2026-08-11 |
| **Significance** | 1/5 |
| **Breaking** | no |
| **Categories** | — |

> **Note:** this classification is below our confidence threshold and is pending human review.

## What changed


- Pipeline tag: image-text-to-text
- Library: transformers
- Tags: judge-model, image-text-to-text, text-to-image, evaluation, benchmark, conversational, en, zh
- Base model: Qwen/Qwen3.6-27B (fine-tuned)
- License: Apache 2.0
- Arxiv paper reference: 2605.28091


## Sources

- [hf_model](https://huggingface.co/Qwen/Qwen-Image-Bench) — retrieved 2026-08-11


## Community

_No curated reactions recorded. Facts and community takes are kept in separate layers
and never blended._

---
Canonical: https://changelogs.info/qwen/qwen-publishes-qwen-image-bench-evaluation-model
Entity: https://changelogs.info/qwen
Event ID: `evt_2026-05-21_qwen_qwen-qwen-image-bench`
Licence: event synthesis © changelogs.info, CC BY 4.0. Linked sources belong to their vendors.