Abstract glowing figure with distorted shapes and glitch effects, embodying synthesis, phase twist, and pain in a dark environment.
A futuristic cybernetic robot wearing a glossy black bodysuit with gold claws and a cyberpunk demon helmet, standing in a dynamic pose against a cityscape background.
Futuristic cybernetic warrior with a sleek, metallic body holding two glowing swords. AI-generated image using Stable Diffusion.

Recommended Prompts

exosuit,tactical gear,futuristic,cyberpunk

exosuit, tactical, advanced gear, futuristic, hero pose, intricate, highly detailed, vignette, high budget, film grain, bokeh, cinemascope, moody, epic, gorgeous, grainy

Recommended Negative Prompts

anime, graphic, text, painting, crayon, graphite, abstract, deformed, glitch, realism, photorealistic, ugly, mutated, cartoon, disfigured, bad quality, low quality, blurry, worst quality, (hands:1.3), (hair:1.2), lowres, meme

Recommended Parameters

samplers

dpmpp_3m_sde_gpu_karras

steps

40

cfg

4

resolution

1x1

vae

VAE_fixFP16ErrorsSDXLLowerMemoryUse_v10

Tips

Recommended weight for version 1.0 is about 0.8 (range 0.5 - 1.0)

For version 1.5, use weight 1.0 (works between 0.5 and 1.3-1.5)

No specific triggers required, but keywords like exosuit, tactical gear, futuristic, cyberpunk work best

Use random prompts with a few keywords added for image creation

Permission granted to use Civitai generation services

Version Highlights

I tried to refine the previous version slightly. This one should allow for even more consistent features for the cyborg Lora. Hope you enjoy this version as well.

You can now easily use a strength 1.0 with this Lora.
I made sure not to fry it this time!

I added some new pics including the prompt, so it should be easier for you to use the lora.

This will for sure also not be the last version, as I am looking for more material to train the Lora on to improve its quality and versatility.

Again, there are no real trigger words, but words like e.g. exosuit, tactical gear, futuristic, cyberpunk, work best.

I am finally releasing one of my costume Loras to the public. Enjoy and play around with it.
Recommended weight for v1.0 is about 0.8 but it should work fine between 0.5 - 1.0.
For v1.5 you can easily use a weight of 1.0 now, and it should work fine between 0.5 up to 1.3-1.5

All images are created with random prompts and just a few keywords added. There are no specific triggers required.

There are no real trigger words, but words like e.g. exosuit, tactical gear, futuristic, cyberpunk, work best.

Have fun and post a picture or two!

Added some new pics to v1.5 with the prompt included!

Also added the permission to use the Civitai generation services.

Contributor

Previous
Qwen-Image-Edit FP8 & GGUF - Q3_K_M.gguf
Next
AnyLoRA - Checkpoint - bakedVae (blessed) fp16 NOT-PRUNED

Model Details

Model type

LORA

Base model

SDXL 1.0

Model version

v1.5

Model hash

0dbe8e4249

Discussion

Please log in to leave a comment.

Images by CyborgMechExosuit - v1.5

character Images

cyberpunk Images

cyborg Images

mech Images

tactical Images