Krea2 Depth Map Spatial Control Workflow

This workflow is for Krea2 depth-map spatial control. It uses a depth map to guide the image structure, pose, perspective, and spatial relationship before generating the final Krea2 image. The goal is to make Krea2 output less random and more controllable when the creator wants to preserve body angle, camera distance, composition, or foreground-background depth.

The workflow uses krea2_turbo_bf16.safetensors as the main model, qwen3vl_4b_fp8_scaled.safetensors as the Krea2 text encoder, and qwen_image_HDR_vae_fp32_comfy.safetensors as the VAE. The control route uses DepthAnythingV2Preprocessor with depth_anything_v2_vitl.pth, then encodes the depth result through Krea2ControlImageEncode and applies it with Krea2ControlApply. The workflow also loads depth-control-lora.safetensors through Krea2ControlLoRALoader at strength 1.

This makes the workflow useful for controlled portraits, pose-guided image generation, spatially coherent character shots, scene reconstruction, and composition transfer. Instead of only describing the image in text, you can use a depth reference to control where the subject is, how the body is angled, how the camera frames the scene, and how the image depth should behave.

The workflow includes two depth-control branches, making it useful for comparing different depth-guided outputs. The sampling route uses 8 steps, CFG 1, euler sampler, simple scheduler, and full denoise.

Main features:

• Krea2 depth-map spatial control

• DepthAnythingV2 preprocessing

• depth_anything_v2_vitl.pth support

• depth-control-lora.safetensors support

• Krea2ControlImageEncode route

• Krea2ControlApply route

• Krea2 Turbo BF16 model

• Qwen3-VL Krea2 text encoder

• Qwen Image HDR VAE

• 8-step euler sampling

• Suitable for pose control, camera control, and depth-guided composition

Suggested workflow:

Use a reference image with a clear pose and readable depth. The depth map should guide structure, not replace the prompt. In the prompt, describe the new subject, style, lighting, and atmosphere while preserving the spatial intention from the depth input. If the output ignores the depth too much, use a clearer reference image or simplify the prompt. If the result feels too rigid, choose a softer depth reference.

⚙️ RunningHub Workflow

Try the workflow online right now — no installation required.
👉 Workflow: https://www.runninghub.ai/post/2074545419065053185?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/2074545419065053185?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。