One photo of a cat,
one identity found

Add cat identity recognition to your application.

Enroll photos for each cat in a project, then search any new photo against that project. Enrollment, multi-cat detection and ranked results come from one OpenAPI.

Project-level isolationNo per-project trainingREST API
Green glass cat identity core surrounded by fingerprint and profile modules
Core capabilitiesFace and body featuresSeveral cats, one photoRanked by similarityRegistry per project

See cat face recognition in action

Explore MeowID in this English-language video.

Watch on YouTube

From “a cat” to “this cat”

Detection, enrollment and identity search form one workflow. MeowID handles recognition; your application sets the business rules.

01

Face first, body as fallback

A clear face drives the comparison; when the face is hidden, body markings are used instead.

Explore recognition
02

A profile for every cat

Manage names, custom fields and enrollment photos. Add images and set a primary photo as your registry grows.

Explore enrollment
03

Every cat in one photo

Query images are checked for several cats, each returning candidates, ranked results and its own crop.

Read multi-cat results
04

Projects stay separate

Keep project data separate. Configure API key scopes, expiry, status, revocation and IP allowlists.

Read about authentication
05

Enrollment you can follow

Follow asynchronous enrollment, select the intended cat from a multi-cat image, and retry failed tasks when source data remains.

Explore task workflows
06

You set the threshold

Choose how many results to return and where the similarity cutoff sits. Uncertain matches return an unknown state for manual review.

Understand the response

From creating a project to the first match

Read the integration workflow
  1. 01

    Create a project and key

    Create a project in the console and configure a project-specific API key with the permissions you need.

  2. 02

    Build a cat profile

    Create an identity, upload enrollment images and wait for background feature extraction.

  3. 03

    Search with a photo

    Call the search endpoint to receive candidate identities, similarity scores and unknown states for each cat.

No model training, no inference servers

Call it with plain HTTP requests — requests or fetch is enough. Detection and feature extraction run on the platform.

  • Bearer authentication with scoped keys
  • cURL, Python and Postman integration examples
  • Enrollment status and results stay queryable
Read developer documentation Illustrated API integration guide
Recognition requestcURL
# Set your API URL, project ID and server-side key
curl --request POST \
  "$MEOWID_API_BASE/v1/projects/{project_id}/search?top_k=5" \
  --header "Authorization: Bearer $MEOWID_API_KEY" \
  --form "[email protected];type=image/jpeg"
Query after enrollment completesAPI reference

Where MeowID fits

Product directions built around cat identity.

Campus and community registries

Keep photos taken at different times in one profile, supporting observation, volunteer work and cat records.

Rescue and follow-up records

Link rescue, care and follow-up notes to a cat identity. Candidates narrow the search; staff confirm the match.

Pet apps and services

Add cat recognition to a mini app, website or business system, making photo upload, identity search and registry management part of your product.

Questions before you integrate

What is MeowID? Does it identify cat breeds?

MeowID is a cat identity platform for developers. It links enrollment images to cat profiles and searches for individual identities in new photos. The current API answers whether this is the same cat; it does not classify breeds.

Do I need to enroll a cat before recognition?

Yes. Create a cat identity in a project, upload enrollment images and wait for feature extraction to finish. Queries search that project's enrolled registry. If no reliable match meets the threshold, the result is marked unknown.

What if the face is partly hidden, or there are several cats?

The platform uses face or body features according to detection results. Queries can return candidates for each detected cat. Multi-cat enrollment may require selecting the intended cat first. Lighting, sharpness and occlusion affect results.

Is a similarity score an accuracy percentage?

No. Similarity is used for candidate ranking and threshold decisions; it is not the probability that an identity is correct or the platform's accuracy. Detection scores also differ from identity similarity. Validate thresholds on your own images and review important results.

Do I need to log in to read the documentation?

No. Documentation is available to read directly. Enrollment, recognition and data queries require valid credentials and permissions. Keep API keys on your server. Project creation, key management and billing require a console login.

How is cat recognition billed?

Enrollment and recognition are charged for cat embedding extraction, against the key owner's balance. Check the console for current prices, grants and usage.

Contact us

Reach us by email about integration, product feedback or partnerships.

Email[email protected]Send an email

When reporting an API issue, include the endpoint, HTTP status and request ID. Never send API keys or access tokens by email.

Register your first cat

Start with the documentation

MeowID · Cat Recognition Platform