lkbeauty007-bit/leahavatardemo 🖼️🔢❓📝✓ → 🖼️

▶️ 4 runs 📅 Jan 2026 ⚙️ Cog 0.15.3
lora lora-person text-to-image

Performance

9.2sTypical run time
4Total runs

About

Example Output

Prompt:

"

A hyper-realistic portrait of Leah, created strictly from uploaded reference photos. Do not change body size, proportions, posture, weight, or shape in any way. Preserve Leah’s natural physique exactly as shown in the reference images. No slimming, no reshaping, no enhancement.

Facial features, bone structure, skin tone, and likeness must closely and accurately match the reference photos with no facial alteration.

Leah is wearing professional, modest attire similar to what she wears in real life. Clothing must not alter or sculpt the body.

Hair styled similar to reference photos, softly polished and camera-ready.

Makeup is professionally done and presentable, suitable for filming, education, or podcast content:
– Even, healthy-looking complexion
– Defined eyebrows
– Soft eye makeup (mascara, subtle liner or shadow)
– Natural contouring for dimension (not glam, not heavy)
– Healthy lip color (not pale, not washed out)

Leah should look alive, confident, polished, and well-rested, not flat or emotionless. Expression is calm and approachable with softness in the eyes.

Fingernails should appear neatly groomed and professional.

Framed as a professional talking-head portrait (head and upper torso only).

Background is a luxury medical spa with a clean, modern aesthetic. Subtle blurred elements such as treatment equipment, shelving, or clinic lighting.

Realistic, flattering lighting that enhances natural skin tone. Cinematic realism, shallow depth of field, realistic camera lens effect, true-to-life skin texture, ultra-high realism

"

Output

Example output

Performance Metrics

9.19s Prediction Time
9.23s Total Time
All Input Parameters 
{
  "model": "dev",
  "prompt": "A hyper-realistic portrait of Leah, created strictly from uploaded reference photos. Do not change body size, proportions, posture, weight, or shape in any way. Preserve Leah’s natural physique exactly as shown in the reference images. No slimming, no reshaping, no enhancement.\n\nFacial features, bone structure, skin tone, and likeness must closely and accurately match the reference photos with no facial alteration.\n\nLeah is wearing professional, modest attire similar to what she wears in real life. Clothing must not alter or sculpt the body.\n\nHair styled similar to reference photos, softly polished and camera-ready.\n\nMakeup is professionally done and presentable, suitable for filming, education, or podcast content:\n– Even, healthy-looking complexion\n– Defined eyebrows\n– Soft eye makeup (mascara, subtle liner or shadow)\n– Natural contouring for dimension (not glam, not heavy)\n– Healthy lip color (not pale, not washed out)\n\nLeah should look alive, confident, polished, and well-rested, not flat or emotionless. Expression is calm and approachable with softness in the eyes.\n\nFingernails should appear neatly groomed and professional.\n\nFramed as a professional talking-head portrait (head and upper torso only).\n\nBackground is a luxury medical spa with a clean, modern aesthetic. Subtle blurred elements such as treatment equipment, shelving, or clinic lighting.\n\nRealistic, flattering lighting that enhances natural skin tone. Cinematic realism, shallow depth of field, realistic camera lens effect, true-to-life skin texture, ultra-high realism",
  "go_fast": false,
  "lora_scale": 1,
  "megapixels": "1",
  "num_outputs": 1,
  "aspect_ratio": "9:16",
  "output_format": "webp",
  "guidance_scale": 3,
  "output_quality": 80,
  "prompt_strength": 0.8,
  "extra_lora_scale": 1,
  "num_inference_steps": 28
}
Input Parameters
mask Type: string
Image mask for image inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.
seed Type: integer
Random seed. Set for reproducible generation
image Type: string
Input image for image to image or inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.
model Default: dev
Which model to run inference with. The dev model performs best with around 28 inference steps but the schnell model only needs 4 steps.
width Type: integerRange: 256 - 1440
Width of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation
height Type: integerRange: 256 - 1440
Height of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation
prompt (required) Type: string
Prompt for generated image. If you include the `trigger_word` used in the training process you are more likely to activate the trained object, style, or concept in the resulting image.
go_fast Type: booleanDefault: false
Run faster predictions with model optimized for speed (currently fp8 quantized); disable to run in original bf16
extra_lora Type: string
Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'
lora_scale Type: numberDefault: 1Range: -1 - 3
Determines how strongly the main LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.
megapixels Default: 1
Approximate number of megapixels for generated image
num_outputs Type: integerDefault: 1Range: 1 - 4
Number of outputs to generate
aspect_ratio Default: 1:1
Aspect ratio for the generated image. If custom is selected, uses height and width below & will run in bf16 mode
output_format Default: webp
Format of the output images
guidance_scale Type: numberDefault: 3Range: 0 - 10
Guidance scale for the diffusion process. Lower values can give more realistic images. Good values to try are 2, 2.5, 3 and 3.5
output_quality Type: integerDefault: 80Range: 0 - 100
Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs
prompt_strength Type: numberDefault: 0.8Range: 0 - 1
Prompt strength when using img2img. 1.0 corresponds to full destruction of information in image
extra_lora_scale Type: numberDefault: 1Range: -1 - 3
Determines how strongly the extra LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.
replicate_weights Type: string
Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'
num_inference_steps Type: integerDefault: 28Range: 1 - 50
Number of denoising steps. More steps can give more detailed images, but take longer.
disable_safety_checker Type: booleanDefault: false
Disable safety checker for generated images.
Output Schema

Output

Type: array • Items Type: string • Items Format: uri

Example Execution Logs 
free=23211086348288
Downloading weights
2026-01-16T01:24:35Z | INFO  | [ Initiating ] chunk_size=150M dest=/tmp/tmp4ibz2g7a/weights url=https://replicate.delivery/xezq/qrFCIyFK3zbbBtalAMsq6wEAvGFcPIE5fSgccQK2dSlTD5eVA/trained_model.tar
2026-01-16T01:24:35Z | INFO  | [ Cache Service ] enabled=true scheme=http target=hermes.services.svc.cluster.local
2026-01-16T01:24:35Z | INFO  | [ Cache URL Rewrite ] enabled=true target_url=http://hermes.services.svc.cluster.local/replicate.delivery/xezq/qrFCIyFK3zbbBtalAMsq6wEAvGFcPIE5fSgccQK2dSlTD5eVA/trained_model.tar url=https://replicate.delivery/xezq/qrFCIyFK3zbbBtalAMsq6wEAvGFcPIE5fSgccQK2dSlTD5eVA/trained_model.tar
2026-01-16T01:24:35Z | INFO  | [ Redirect ] redirect_url=http://r8-east4-loras-ric2.cwlota.com/012602f257edd9b116ee2f752281657e8a76214303ee0be8a8296b0184a5594c?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Checksum-Mode=ENABLED&X-Amz-Credential=CWYAFKQUBIDDRADP%2F20260116%2FUS-EAST-04A%2Fs3%2Faws4_request&X-Amz-Date=20260116T012435Z&X-Amz-Expires=600&X-Amz-SignedHeaders=host&x-id=GetObject&X-Amz-Signature=45291ab8cebc3f59c7a9b2eab0b580f5e48404f4d4f42be03a610e8c6db2339a url=http://hermes.services.svc.cluster.local/replicate.delivery/xezq/qrFCIyFK3zbbBtalAMsq6wEAvGFcPIE5fSgccQK2dSlTD5eVA/trained_model.tar
2026-01-16T01:24:36Z | INFO  | [ Complete ] dest=/tmp/tmp4ibz2g7a/weights size="172 MB" total_elapsed=0.549s url=https://replicate.delivery/xezq/qrFCIyFK3zbbBtalAMsq6wEAvGFcPIE5fSgccQK2dSlTD5eVA/trained_model.tar
Downloaded weights in 0.61s
Loaded LoRAs in 1.69s
Using seed: 27208
Prompt: A hyper-realistic portrait of Leah, created strictly from uploaded reference photos. Do not change body size, proportions, posture, weight, or shape in any way. Preserve Leah’s natural physique exactly as shown in the reference images. No slimming, no reshaping, no enhancement.
Facial features, bone structure, skin tone, and likeness must closely and accurately match the reference photos with no facial alteration.
Leah is wearing professional, modest attire similar to what she wears in real life. Clothing must not alter or sculpt the body.
Hair styled similar to reference photos, softly polished and camera-ready.
Makeup is professionally done and presentable, suitable for filming, education, or podcast content:
– Even, healthy-looking complexion
– Defined eyebrows
– Soft eye makeup (mascara, subtle liner or shadow)
– Natural contouring for dimension (not glam, not heavy)
– Healthy lip color (not pale, not washed out)
Leah should look alive, confident, polished, and well-rested, not flat or emotionless. Expression is calm and approachable with softness in the eyes.
Fingernails should appear neatly groomed and professional.
Framed as a professional talking-head portrait (head and upper torso only).
Background is a luxury medical spa with a clean, modern aesthetic. Subtle blurred elements such as treatment equipment, shelving, or clinic lighting.
Realistic, flattering lighting that enhances natural skin tone. Cinematic realism, shallow depth of field, realistic camera lens effect, true-to-life skin texture, ultra-high realism
[!] txt2img mode
  0%|          | 0/28 [00:00<?, ?it/s]
  4%|▎         | 1/28 [00:00<00:06,  3.86it/s]
  7%|▋         | 2/28 [00:00<00:05,  4.37it/s]
 11%|█         | 3/28 [00:00<00:06,  4.13it/s]
 14%|█▍        | 4/28 [00:00<00:05,  4.01it/s]
 18%|█▊        | 5/28 [00:01<00:05,  3.96it/s]
 21%|██▏       | 6/28 [00:01<00:05,  3.93it/s]
 25%|██▌       | 7/28 [00:01<00:05,  3.91it/s]
 29%|██▊       | 8/28 [00:02<00:05,  3.89it/s]
 32%|███▏      | 9/28 [00:02<00:04,  3.88it/s]
 36%|███▌      | 10/28 [00:02<00:04,  3.87it/s]
 39%|███▉      | 11/28 [00:02<00:04,  3.87it/s]
 43%|████▎     | 12/28 [00:03<00:04,  3.86it/s]
 46%|████▋     | 13/28 [00:03<00:03,  3.86it/s]
 50%|█████     | 14/28 [00:03<00:03,  3.86it/s]
 54%|█████▎    | 15/28 [00:03<00:03,  3.86it/s]
 57%|█████▋    | 16/28 [00:04<00:03,  3.86it/s]
 61%|██████    | 17/28 [00:04<00:02,  3.85it/s]
 64%|██████▍   | 18/28 [00:04<00:02,  3.85it/s]
 68%|██████▊   | 19/28 [00:04<00:02,  3.85it/s]
 71%|███████▏  | 20/28 [00:05<00:02,  3.85it/s]
 75%|███████▌  | 21/28 [00:05<00:01,  3.86it/s]
 79%|███████▊  | 22/28 [00:05<00:01,  3.86it/s]
 82%|████████▏ | 23/28 [00:05<00:01,  3.86it/s]
 86%|████████▌ | 24/28 [00:06<00:01,  3.85it/s]
 89%|████████▉ | 25/28 [00:06<00:00,  3.85it/s]
 93%|█████████▎| 26/28 [00:06<00:00,  3.85it/s]
 96%|█████████▋| 27/28 [00:06<00:00,  3.85it/s]
100%|██████████| 28/28 [00:07<00:00,  3.85it/s]
100%|██████████| 28/28 [00:07<00:00,  3.88it/s]
Total safe images: 1 out of 1
Version Details
Version ID
f431148728bdc17b581fdc404de4a9d91ee4dd07e87c2965e25f5878efbae31b
Version Created
January 15, 2026
Run on Replicate →