Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a comfyui workflow for consistent ai characters and ai video
Filip R.
About this gig
Stop fixing faces by hand. I build ComfyUI pipelines that keep your AI character on-model across outfits, poses, lighting and scenes, then produce content at volume.
WHO IT'S FOR
AI influencer creators, agencies, brands with a mascot, and game or story studios that need one character in many images.
WHAT YOU GET
- Dataset preparation and a character LoRA, trained and tested
- A clean, labelled ComfyUI workflow (FLUX or SDXL, IP-Adapter, ControlNet, face detailing)
- Batch generation from a prompt list or spreadsheet
- Image-to-video workflow (Content Engine)
- A consistency report: a fixed 12-prompt drift test that shows where the character holds and where it drifts, so you see proof, not cherry-picks
WHY ME
I run my own multi-lane ComfyUI studio on a local GPU and test every pipeline on real hardware before delivery. You get the JSON workflow, the settings and a guide, not a black box.
BEFORE ORDERING
Message me 3-5 reference images (or a character description) and your GPU or cloud setup. I'll confirm exactly what's achievable.
Safe-for-work projects only.
Delivery style preference
Please inform the freelancer of any preferences or concerns regarding the use of AI tools in the completion and/or delivery of your order.
AI Avatar Design disclosure
I understand that by acquiring AI Avatar Design services, I must abide by any and all disclaimer rules required by applicable laws and the platform(s) where this content will be posted. Read more about Fiverr’s AI Guidelines.
Get to know Filip R.
Filip R.
ComfyUI Expert for Consistent AI Characters and LoRA Training
- FromSlovakia
- Member sinceOct 2026
- Avg. response time3 hours
- Last delivery4 days
Languages
English, Slovak, Czech
I build production ComfyUI pipelines that keep an AI character on-model across poses, outfits, lighting and scenes, then produce content at volume. Consistent characters (LoRA, IP-Adapter, ControlNet, face detailing), LoRA training, batch generation from a spreadsheet, and image-to-video. Every pipeline ships with a consistency report: a fixed 12-prompt drift test, so you see proof, not cherry-picks. I run my own multi-lane ComfyUI studio on a local GPU and test everything before delivery. You get the JSON workflows, settings and a guide.
My Portfolio
FAQ
Will the pipeline run on my own PC?
I check your GPU first. 12 GB VRAM is a practical minimum for most setups, and I adapt the pipeline to the hardware you have, or set it up on a cloud GPU.
What does the consistency report show?
12 fixed test prompts (new angle, outfit, lighting, expression and more), each marked pass, drifting or rebuild. We fix weak spots before final delivery.
Can the character look like a real person?
Only with that person's written consent. Otherwise I build an original character from your references or description.
Do I own the workflow and the images?
Yes. You get the workflow JSON, the LoRA file (Pipeline and Engine) and the images, with full usage rights for your project.
