Krea2 Random Prompt Creation Workflow

This workflow is for Krea2 random prompt creation and three-stage image rendering. It is designed for creators who want to generate new prompt directions quickly instead of manually writing every subject, scene, style, camera, lighting, and atmosphere combination.

The workflow uses CreaPrompt nodes to build randomized prompt combinations. The large CreaPrompt node contains many selectable prompt categories, including people, pose, fashion, background, places, landscapes, artistic style, artists, camera, lighting, visual effects, and quality terms. This makes the workflow useful for prompt exploration, idea generation, and batch creative testing.

After the prompt is created, the image is generated through Krea2 Turbo BF16 with qwen3vl_4b_fp8_scaled.safetensors as the Krea2 text encoder and qwen_image_vae.safetensors as the VAE. The workflow also uses ComfyUI-Krea2T-Enhancer with strength 1, giving the model route stronger response.

The sampling chain uses ClownsharKSampler_Beta with exponential/ddim and beta57. It is not a simple one-pass render. The first stage generates the base latent with 8 steps and full denoise. Then the latent is upscaled by 1.25x, refined in the second stage with 6 steps and denoise 0.45, upscaled again by 1.25x, and finally refined in the third stage with 6 steps and denoise 0.4. This gives the workflow a stronger final render while keeping the creative process fast.

Main features:

• Krea2 random prompt creation workflow

• CreaPrompt random prompt generator

• Large category-based prompt system

• Krea2 Turbo BF16 model

• Qwen3-VL Krea2 text encoder

• Qwen Image VAE

• Krea2T Enhancer support

• ClownsharKSampler_Beta route

• beta57 sampling

• Three-stage refinement chain

• Two latent upscale passes

• 16:9 panorama output

• Suitable for idea exploration, prompt testing, and creative randomization

Suggested workflow:

Use this workflow when you want fast visual ideas without writing every prompt from scratch. Randomize the prompt categories, generate a few candidates, then keep the strongest concept and refine it manually. The random prompt system is best used as a creative starting point. After a good direction appears, improve the prompt with clearer subject, scene, lighting, style, and composition language.

⚙️ RunningHub Workflow

Try the workflow online right now — no installation required.
👉 Workflow: https://www.runninghub.ai/post/2074464834908942338?inviteCode=rh-v1111

If the results meet your expectations, you can later deploy it locally for customization.

🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!

📺 Bilibili Updates (Mainland China & Asia-Pacific)

If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.
📺 Bilibili Video: https://www.bilibili.com/video/BV1nxM56cEta/

☕ Support Me on Ko-fi

If you find my content helpful and want to support future creations, you can buy me a coffee ☕.
Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.
👉 Ko-fi: https://ko-fi.com/aiksk

💼 Business Contact

For collaboration or inquiries, please contact aiksk95 on WeChat.

⚙️打开下方链接即可在线体验，无需安装。
👉 工作流： https://www.runninghub.ai/post/2074464834908942338?inviteCode=rh-v1111
如果觉得效果理想，你也可以在本地进行自定义部署。

🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！

📺 Bilibili 更新（中国大陆及南亚太地区）

如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。
📺 B站视频： https://www.bilibili.com/video/BV1nxM56cEta/

我会在 夸克网盘 持续更新模型资源：
👉 https://pan.quark.cn/s/20c6f6f8d87b
这些资源主要面向本地用户，方便进行创作与学习。