videobgremover-pro: Background Removal with Text Prompts

Text prompts let you use natural language to specify exactly what to keep in the video:

Example prompts:

• "person wearing red jacket"
• "golden retriever dog"
• "blue sports car"
• "wooden dining table"
• "white cat with black spots"

Instead of automatic detection, the AI uses your description to identify and isolate the specific subject, removing everything else. This gives you precise control over what stays and what gets removed.

videobgremover-pro costs 180 credits per minute (3.0 credits/sec), which is 3x the cost of the original model:

• 30-second video: 90 credits (vs 30 credits for original)
• 2-minute video: 360 credits (vs 120 credits for original)
• 5-minute video: 900 credits (vs 300 credits for original)

The premium price reflects the advanced text prompt AI and ability to handle complex object segmentation that other models can't do.

Use videobgremover-pro when:

• You need text prompt control to specify exactly what to keep
• Multiple subjects exist and automatic detection picks the wrong one
• Complex objects that automatic models struggle with (specific animals, products, etc.)
• Precise segmentation is critical and automatic methods aren't working

For most videos, videobgremover-original, videobgremover-light, or videobgremover-human will work great and cost less. Use Pro only when you need its unique text prompt capability.

Good prompts are:

• Specific: "golden retriever" instead of "dog"
• Descriptive: "person wearing red jacket" instead of "person"
• Clear: "blue sports car" instead of "the car"

Example prompts that work well:

• Animals: "black labrador", "orange tabby cat", "white horse"
• People: "person in blue shirt", "woman with glasses", "child with backpack"
• Products: "red sneakers", "wooden chair", "silver laptop"
• Vehicles: "white SUV", "blue sports car", "yellow taxi"

The more specific your description, the better the AI can identify your subject.

Yes, but it wouldn't be cost-effective!

videobgremover-pro can work in automatic mode (without text prompts), but you'd be paying 3x the price for the same automatic detection available in other models.

Better options for automatic detection:

• videobgremover-original - Highest quality, 60 credits/min
• videobgremover-light - Fast & cheap, 45 credits/min
• videobgremover-human - Fastest for people, 45 credits/min

Only use videobgremover-pro when you specifically need text prompt capabilities.

videobgremover-pro excels at isolating specific objects from multi-subject scenes:

Example scenario: Video has 3 people and 2 dogs. You want only the golden retriever.

Without Pro (automatic):

• Might keep all subjects
• Might pick the wrong subject
• No way to specify which one to keep

With Pro (text prompt):

• Prompt: "golden retriever"
• AI identifies only that specific dog
• Removes everything else including the other subjects

This precision control is what makes videobgremover-pro unique.

Yes, videobgremover-pro is the slowest model because it uses more advanced AI to process your text prompts:

• Processing time: Similar to videobgremover-original (standard speed)
• Why slower: Text understanding + advanced segmentation requires more compute
• Trade-off: Precise control and complex object handling justify the extra time

If speed is critical and you don't need text prompts, use videobgremover-light (2x faster) or videobgremover-human (fastest).

Yes! The Python and Node.js SDKs fully support videobgremover-pro with text prompts:

await client.startJob(job.id, {
  model: 'videobgremover-pro',
  prompt: { mode: 'text', text: 'golden retriever dog' },
  background: { type: 'transparent', transparent_format: 'webm_vp9' }
})

View API Documentation for complete examples and implementation details.