VicSeeVicSee

Character Creation

Train a consistent character identity from photos (FLUX LoRA). Upload 10-30 images, get a character ID, then generate unlimited consistent images via API. 500 credits per training, 15 credits per image.

New: Train a consistent character identity from photos. Upload 10-30 images of a person, train a FLUX LoRA, then generate unlimited images of that person in any scene, outfit, or style.

Pricing

ActionCreditsPrice (Pro Yearly)Price (Pro Monthly)
Train character500$3.00$4.00
Generate image15 per image$0.09$0.12

Training credits are deducted upfront when you create a character. Refunded automatically if training fails.

Workflow

  1. Create — Upload 10-30 photos of a person. Returns character_id immediately.
  2. Poll — Check training status (takes ~2-3 minutes).
  3. Generate — Once ready, generate unlimited consistent images.

Create Character

Train a new character identity from photos.

POST https://vicsee.com/api/v1/characters/create

See Authentication for API key setup.

Parameters

ParameterTypeRequiredDescription
image_urlsstring[]YesArray of 10-30 public image URLs (JPEG, PNG, WebP). More images = better consistency.
namestringNoDisplay name for the character. Default: "Character"
trigger_wordstringNoUnique keyword to activate the character in prompts. Auto-generated if not provided.

Example Request

curl -X POST https://vicsee.com/api/v1/characters/create \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Sarah",
    "image_urls": [
      "https://example.com/photo1.jpg",
      "https://example.com/photo2.jpg",
      "https://example.com/photo3.jpg",
      "https://example.com/photo4.jpg",
      "https://example.com/photo5.jpg",
      "https://example.com/photo6.jpg",
      "https://example.com/photo7.jpg",
      "https://example.com/photo8.jpg",
      "https://example.com/photo9.jpg",
      "https://example.com/photo10.jpg"
    ]
  }'

Response

{
  "success": true,
  "data": {
    "character_id": "abc123-def456",
    "name": "Sarah",
    "trigger_word": "CHAR1N2K3M4",
    "status": "training",
    "image_count": 10,
    "credits_used": 500,
    "created_at": "2026-03-01T12:00:00.000Z"
  }
}

Training takes approximately 2-3 minutes. Use the status endpoint to check when it's ready.

Check Status

Get character details and training status.

GET https://vicsee.com/api/v1/characters/{character_id}

Response (Training)

{
  "success": true,
  "data": {
    "character_id": "abc123-def456",
    "name": "Sarah",
    "trigger_word": "CHAR1N2K3M4",
    "status": "training",
    "image_count": 10,
    "created_at": "2026-03-01T12:00:00.000Z"
  }
}

Response (Ready)

{
  "success": true,
  "data": {
    "character_id": "abc123-def456",
    "name": "Sarah",
    "trigger_word": "CHAR1N2K3M4",
    "status": "ready",
    "image_count": 10,
    "created_at": "2026-03-01T12:00:00.000Z"
  }
}

Status Values

StatusDescription
trainingLoRA training in progress (~2-3 minutes)
readyCharacter trained and ready for generation
failedTraining failed (credits refunded automatically)

Generate Images

Generate consistent images of a trained character.

POST https://vicsee.com/api/v1/characters/{character_id}/generate

Parameters

ParameterTypeRequiredDescription
promptstringYesText description of the scene. The trigger word is auto-injected.
aspect_ratiostringNoImage aspect ratio. Default: "1:1".
num_outputsintegerNoNumber of images (1-4). Default: 1.

Aspect Ratios

1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3

Example Request

curl -X POST https://vicsee.com/api/v1/characters/abc123-def456/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "professional headshot in a modern office, wearing a navy blazer",
    "aspect_ratio": "3:4",
    "num_outputs": 2
  }'

Response

{
  "success": true,
  "data": {
    "id": "task_xyz789",
    "character_id": "abc123-def456",
    "status": "completed",
    "output": {
      "images": [
        { "url": "https://cdn.vicsee.com/results/dew/user-id/img1.png" },
        { "url": "https://cdn.vicsee.com/results/dew/user-id/img2.png" }
      ]
    },
    "credits_used": 30,
    "credits_remaining": 470,
    "created_at": "2026-03-01T12:05:00.000Z"
  }
}

Generation takes approximately 30 seconds per image. Output URLs are hosted on VicSee CDN (available for 7 days).

Complete Pipeline Example

Train a character and generate marketing photos:

# Step 1: Create character
CHAR_ID=$(curl -s -X POST https://vicsee.com/api/v1/characters/create \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Brand Ambassador",
    "image_urls": ["https://...", "https://...", "...10-30 URLs..."]
  }' | jq -r '.data.character_id')

echo "Character ID: $CHAR_ID"

# Step 2: Poll until ready (typically 2-3 minutes)
while true; do
  STATUS=$(curl -s https://vicsee.com/api/v1/characters/$CHAR_ID \
    -H "Authorization: Bearer YOUR_API_KEY" | jq -r '.data.status')
  echo "Status: $STATUS"
  [ "$STATUS" = "ready" ] && break
  [ "$STATUS" = "failed" ] && echo "Training failed" && exit 1
  sleep 10
done

# Step 3: Generate marketing photos
curl -X POST https://vicsee.com/api/v1/characters/$CHAR_ID/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "professional photoshoot, studio lighting, white background",
    "aspect_ratio": "3:4",
    "num_outputs": 4
  }'

Photo Guidelines

For best results:

  • Minimum 10 photos, 20+ recommended for best quality
  • Include variety: different angles, lighting, expressions, backgrounds
  • Clear, well-lit photos of the person's face
  • Avoid heavy makeup, sunglasses, or face-obscuring accessories
  • Higher resolution = better results (minimum 512x512)
  • Consistent appearance: same hairstyle/hair color across photos

Limits

ConstraintValue
Minimum images per character10
Maximum images per character30
Maximum image size10 MB per image
Supported image formatsJPEG, PNG, WebP
Training time~2-3 minutes
Generation time~30 seconds per image
Max images per generation4
Output URL availability7 days

Errors

CodeHTTPDescription
TOO_FEW_IMAGES400Fewer than 10 image URLs provided
TOO_MANY_IMAGES400More than 30 image URLs provided
INVALID_IMAGE_URL400Image URL is not a valid HTTP(S) URL
MISSING_PROMPT400No prompt provided for generation
INVALID_ASPECT_RATIO400Unsupported aspect ratio
CHARACTER_NOT_READY422Character is still training or failed
IMAGE_DOWNLOAD_FAILED422Could not download one or more images
INSUFFICIENT_CREDITS402Not enough credits
NOT_FOUND404Character not found or not owned by user
TRAINING_SUBMIT_FAILED500Failed to start training on Replicate
GENERATION_FAILED500Image generation failed or timed out

Generate Music

Generate AI music from text prompts using Suno V5. Each generation returns 2 song variations with cover art. 15 credits per generation. Simple mode or custom mode with lyrics, style, and title control.

Lipsync

Lip-sync video characters to an audio track. Modify lip movements in AI-generated or real video to match voiceover narration. Tiered pricing from 50 credits. Max 30s.

On this page

Pricing
Workflow
Create Character
Parameters
Example Request
Response
Check Status
Response (Training)
Response (Ready)
Status Values
Generate Images
Parameters
Aspect Ratios
Example Request
Response
Complete Pipeline Example
Photo Guidelines
Limits
Errors
Related