Boogu-Image-0.1 and Krea 2 Shake Up AI Gravure Creation

AI-generated gravure image creation options are expanding. The arrival of Boogu-Image-0.1 and Krea 2 raises the possibility that dependence on existing models like Z-Image may change.

📑Table of Contents

Main Features and Variants of Boogu-Image-0.1
Background and Technical Advantages of Krea 2
Performance Comparison with Z-Image (Speed, Quality, Editing)
Practical Usage Examples in Gravure Production
Frequently Asked Questions (FAQ)
Summary

Main Features and Variants of Boogu-Image-0.1

Boogu-Image-0.1 is an open-source model with 10B parameters under the Apache 2.0 license. Its ability to handle image generation and editing in a unified way is a key feature.

There are three main variants. The Base version excels in diversity and controllability, with strong text rendering at 2K resolution. The Turbo version is accelerated through 4-step distillation and supports LoRA rank-128, enabling high-quality image generation in 3–4 steps. The Edit version is excellent at texture preservation and is suited for instruction-driven editing of concrete objects, sketches, and skin.

On Hugging Face, it is published as Boogu/Boogu-Image-0.1-Edit and Comfy-Org/Boogu-Image, with ComfyUI workflows also provided at docs.comfy.org. It recorded 53.58 points on Qwen-Image-Bench, outperforming competitors in text rendering and editing fidelity.

Sources: Hugging Face Boogu-Image-0.1-Edit, ComfyUI docs (as of June 2026)

Background and Technical Advantages of Krea 2

Krea 2 is a foundational image model announced on May 12, 2026, by krea.ai. The fact that the Krea team built it from scratch is a distinguishing point, with a focus on aesthetics, style transfer, and creative control.

On June 3, 2026, Krea 2 Turbo was additionally announced, capable of generating high-quality images in 2 seconds. It offers style reference, mood boards, and compatibility with LoRA. Raw and Turbo open-weights versions are available on Hugging Face under a custom license that requires technical safeguards. Enterprise use (50+ seats) is paid.

The official krea.ai site emphasizes 4K native output and a minimalist UI. VentureBeat reported on June 23, 2026, “Enterprise-grade AI image generation in 2 seconds is here,” highlighting the significance of open-weights release.

Sources: krea.ai blog, VentureBeat (as of June 2026)

Performance Comparison with Z-Image (Speed, Quality, Editing)

Comparing Boogu-Image-0.1 and Krea 2 with Z-Image clarifies differences in speed, quality, and editing.

Item	Boogu-Image-0.1 Turbo	Krea 2 Turbo	Z-Image (reference)
Generation speed	3-4 steps (fast)	2 seconds	Closed model baseline
License	Apache 2.0 (open)	Custom (open-weights)	Closed
Editing features	Edit version: instruction-driven, texture preservation	Style reference, LoRA compatible	Basic editing
Text rendering	Strong at 2K resolution	Style transfer emphasis	Standard
Benchmark	Qwen-Image-Bench 53.58	Official emphasis on aesthetics & speed	Not disclosed

Boogu-Image-0.1 offers high freedom for local execution and customization, but requires building a ComfyUI environment. Krea 2 Turbo balances speed and open-weights, though license confirmation is essential for commercial use. Z-Image, being closed, has lower customizability and stronger API dependence.

Sources: Official announcements and VentureBeat (as of June 2026)

Practical Usage Examples in Gravure Production

From a gravure photographer’s perspective, the following workflows are possible with these models.

First, use the Boogu-Image-0.1 Edit version to correct skin or costume textures in existing photos while keeping them intact. A prompt example such as “Keep skin texture natural while softening the lighting” yields high-fidelity editing results.

Krea 2 Turbo can quickly reproduce the taste of a specific photographer or magazine by leveraging style reference. The 2-second generation makes it easy to try numerous variations rapidly. Combining it with LoRA also improves reproducibility of specific models and poses.

As a difference from Z-Image, Boogu and Krea can be operated locally or with open-weights, giving them advantages in data security and cost. On the other hand, closed models like Z-Image offer high immediacy and stability, but editing freedom is limited.

In practice, building a Boogu-Image-0.1 workflow in ComfyUI and using Krea 2 as a supplement is effective. Adding final checks and fine adjustments by a gravure photographer after generation ensures commercial-level quality.

Frequently Asked Questions (FAQ)

Q: Will Boogu-Image-0.1 and Krea 2 become complete alternatives to Z-Image?

They will not be complete alternatives. They have advantages in speed and openness, but may not match Z-Image’s stable API quality and immediacy in some cases. Using them together depending on the use case is realistic.

Q: Is commercial use possible?

Boogu-Image-0.1 is commercially usable under Apache 2.0. Krea 2 requires a paid plan for enterprises with 50+ seats under its custom license. Please check the API terms for Z-Image.

Q: Can it be used outside the ComfyUI environment?

Boogu-Image-0.1 has rich official ComfyUI workflows, but Hugging Face models can also be loaded in other UIs or scripts. Krea 2 supports the official UI and Hugging Face.

Q: How is the text rendering quality?

The Boogu-Image-0.1 Base version is strong at text rendering in 2K resolution and received high marks on Qwen-Image-Bench. It is suitable for inserting text in gravure work.

Q: Is the hurdle for local execution high?

The 10B parameter size requires GPU memory. The Turbo version has fewer steps, and operation reports exist even on Mac or mid-range GPUs. Refer to the ComfyUI installation guide.

Q: Is the 2-second generation of Krea 2 Turbo real?

It has been confirmed in official announcements and VentureBeat reporting. The design maintains high speed even when using style reference or LoRA.

Related articles:

Summary

Boogu-Image-0.1 and Krea 2 are rising open-source/open-weights contenders that could challenge Z-Image’s stronghold. In gravure production, they offer high practicality in terms of editing freedom and speed, expanding options for commercial use.

Going forward, we plan to actually try these models in ComfyUI and verify the differences with Z-Image. We encourage readers to access the official demos and Hugging Face to consider using them as well.

Sources: krea.ai, Hugging Face, VentureBeat, ComfyUI docs (as of June 2026)

Related new article:

Hiroyuki: “I Escaped the Hardship” — AI Strengthening 40-50s While Squeezing Out Young Engineers – This published update adds current operational context for Boogu-Image-0.1 and Krea 2 Shake Up AI Gravure Creation — Challenging Z-Image.

Author

krona23

Over 20 years in the IT industry, serving as Division Head and CTO at multiple companies running large-scale web services in Japan. Experienced across Windows, iOS, Android, and web development. Currently focused on AI-native transformation. At DevGENT, sharing practical guides on AI code editors, automation tools, and LLMs in three languages.

DevGENT about →

📚 Read Next