VehicoreVehicore
Get started free
Services
Image Classification (CLIP)

Tag what’s in the image with flexible, label-aware search.

Classify images using CLIP-style embeddings with support for custom labels and URL-based inputs.

Use this service See all features

Best for

Common places this service fits well.

  • Search & discovery
  • Dataset labeling
  • Content filtering

Inputs

What you provide.

  • Uploaded image or image URL
  • Optional: custom label set
  • Optional: top-k limit

Outputs

What you get back.

  • Top labels with confidence
  • Embeddings or similarity scores (where enabled)

Notes

Implementation details and product intent.

  • Supports both “zero-shot” style usage and explicit label lists.

Try image classification (CLIP)

Upload an image and get top labels using CLIP.

No file selected

Max size 8 MB. For production, use your own storage or URLs.

Limited demo. For higher volume and auth, sign up and use your API key.

Output
JSON viewer and cURL you can copy.
// Run the demo to see a live
// JSON response here.
// For full docs, sign up and
// open the API reference.

Part of a pipeline?

Classification filters content before more expensive services run.

  1. Photo input
  2. 1Image classification
  3. 2Side recognition
  4. 3Damage
Learn about pipelines
Explore

Related services

Most teams start with one capability, then add the ones that remove friction next.

OCR

Read what’s on the car, not what the camera wished it saw.

Extract text from photos (dash, documents, stickers, part numbers, plates/VIN) with layout-aware results.
Open page

Damage recognition

Turn “looks fine” into measurable findings.

Detect and highlight visible damage areas, returning locations and severity signals you can review.
Open page

Car Side Recognition

Know which side you’re looking at, and where the car actually is.

Classify the vehicle side/orientation and return stable geometry for downstream tasks (cropping, alignment, QA).
Open page