AI Character & Video Specialist – LoRA / Consistent AI Model

We are looking for an experienced AI character / AI video specialist who can work with an existing AI model identity and maintain a highly consistent character across images and video.

This is not a basic AI video generation job.

We specifically need someone who understands LoRA-based character consistency, reference workflows, identity preservation and AI video production.

We already have an AI model/character and may provide an existing LoRA file. You must be comfortable working with an existing LoRA rather than requiring us to rebuild the character from scratch.

The goal is to create premium, highly realistic social-media and campaign content where the same model remains recognisable and consistent across:

different outfits
different locations
different poses
close-ups
full-body shots
lifestyle scenes
cinematic scenes
multiple videos

Facial identity must remain stable. We do not want a character whose face changes significantly between generations.

What we need

You should be able to:

Work with an existing LoRA file
Maintain strong facial and character consistency
Generate photorealistic images using the existing character
Create consistent video clips using the same AI model
Preserve identity across different camera angles and lighting conditions
Create natural body movement and facial movement
Maintain consistent hair, facial structure and key character features
Change wardrobe and environment without losing the character identity
Produce realistic luxury/lifestyle/social-media content
Create vertical 9:16 video suitable for Instagram Reels and TikTok
Upscale and polish final outputs where necessary
Preferred technical experience

We are particularly interested in people experienced with workflows involving tools such as:

ComfyUI
Stable Diffusion / SDXL
FLUX
LoRA
Kohya or similar LoRA workflows
ControlNet
IP-Adapter / FaceID / reference-based identity workflows
PuLID / InstantID or similar identity-preservation methods
Kling
Runway
Veo
Hailuo / MiniMax
Luma
Wan
other advanced image-to-video / character-consistency workflows

You do not need to use every tool listed above. We care about the final quality and your understanding of character consistency.

Important

We already have character assets and may supply:

existing LoRA file
reference images
character/model images
visual direction
example environments
clothing references

You must be able to inspect the existing material and tell us whether you can work directly with it.

We do not want someone who automatically says that a new LoRA must be trained before even reviewing the existing file.

If the existing LoRA genuinely needs improvement, you can recommend that after testing it.

Paid test

We will begin with a small paid test.

For the test, we may ask you to create:

2–3 consistent still images using our existing AI character / LoRA
1 short video clip using the same character

The purpose of the test is to evaluate:

facial consistency
realism
character identity preservation
motion quality
clothing/environment control
ability to work with our existing LoRA

If successful, there is potential for ongoing work creating regular AI model content.

When applying, please answer ALL of the following:
Have you personally worked with LoRA files for a specific human character/model?
Can you work with an existing LoRA file supplied by us?
Which base models do you normally use with LoRAs?
For example: FLUX, SDXL, Stable Diffusion 1.5, etc.
Which workflow do you use for maintaining the same face and identity across multiple images?
Which workflow/tools do you use to maintain the same character in video?
Please send 3–5 examples showing the SAME AI character in different images or videos.

We are much more interested in consistency than random beautiful AI images.

Please clearly state which parts of the examples you personally created.
Can you maintain the same character while changing:
clothing
location
pose
lighting
camera angle
Can you create both:
realistic still images
realistic video clips
Which tools do you personally use?
Do you work locally / through ComfyUI or mainly through web-based AI platforms?
Would any part of the work be outsourced?
What is your fixed price for a paid test consisting of:
2–3 still images
1 short AI video clip
What would you normally charge for ongoing production?
How quickly could you complete the initial test?
Please start your proposal with:

LORA CONSISTENCY

so we know you have read the full brief.

Please do not send a generic AI portfolio.

We specifically want to see the same character reproduced consistently across multiple outputs.

If your portfolio only shows different AI models in every image, this is not the right project.