generated: '2026-07-19'
method: searched
source: https://docs.claid.ai/
docs: https://docs.claid.ai/
api: Claid API
summary: >-
  The Claid API is a declarative JSON pipeline API. A request describes the input image, the
  operations to apply and the desired output; the service returns the processed result. Every
  operation is authenticated with a scoped bearer API key over HTTPS. Work is executed synchronously,
  asynchronously by task id, or in batch, and completion can be pushed to an HMAC-signed webhook.
authentication:
  style: bearer-api-key
  header: Authorization
  format: 'Bearer {YOUR_API_KEY}'
  scoped: true
  docs: https://docs.claid.ai/authentication
  artifact: authentication/lets-enhance-authentication.yml
versioning:
  scheme: uri-path
  current: v1
  base_url: https://api.claid.ai/v1/
  notes: >-
    All published endpoints sit under the /v1/ path prefix. Claid publishes no dated-version or
    version-header scheme and no public version-deprecation policy.
request:
  media_type: application/json
  shape: >-
    A declarative envelope with `input` (a URL to the source image), `operations` (the editing
    pipeline to apply) and `output` (result format and destination). Alternative ingestion is a
    direct multipart upload via POST /v1/image/edit/upload.
  transport: HTTPS only
response:
  media_type: application/json
  envelope: >-
    Successful responses wrap the payload in a top-level `data` object. Image editing responses
    report `input` and `output` metadata (ext, mime, format, width, height, mps) and a `tmp_url`
    for the produced asset.
  example: |-
    {
      "data": {
        "input":  {"ext": "jpeg", "mps": 4.1472, "mime": "image/jpeg", "format": "JPEG", "width": 2880, "height": 1440},
        "output": {"ext": "jpeg", "mps": 0.64,   "mime": "image/jpeg", "format": "JPEG", "width": 800,  "height": 800, "tmp_url": "....."}
      }
    }
idempotency:
  supported: false
  notes: >-
    Claid documents no idempotency key, no request-replay semantics and no Idempotency-Key parameter
    in any published endpoint definition. Re-posting an image processing request creates a new job
    and consumes new API call credits. Callers that need at-most-once semantics must deduplicate on
    their own side, keyed on the task id returned by the async endpoints.
async:
  model: submit-then-poll, with optional webhook push
  submit:
  - POST /v1/image/edit/async
  - POST /v1/image/edit/batch
  - POST /v1/image/ai-edit
  - POST /v1/image/ai-fashion-models
  - POST /v1/video/generate
  accepted_response:
    fields:
    - name: id
      type: integer
      description: Task ID.
    - name: status
      type: string
      description: 'Processing status. For a valid submission this is always: ACCEPTED.'
    - name: result_url
      type: string
      description: >-
        URL of the GET endpoint used to retrieve the result and to check status while processing.
    - name: created_at
      type: string
      description: Exact time the processing was requested.
    - name: request
      type: object
      description: The body of the request that was used for processing.
  status_values:
  - ACCEPTED
  - PROCESSING
  - DONE
  poll:
  - GET /v1/image/edit/async/{task_id}
  - GET /v1/image/edit/batch/{task_id}
  - GET /v1/image/ai-edit/{ai_edit_id}
  - GET /v1/image/ai-fashion-models/{processing_request_id}
  - GET /v1/video/generate/{animation_id}
  docs: https://docs.claid.ai/image-editing-api/async-api-reference
pagination:
  supported: false
  notes: >-
    No published endpoint declares pagination parameters. The only collection endpoints are
    GET /v1/storage/storages and GET /v1/storage/storage-types, which return the full set.
request_tracing:
  header: x-request-id
  direction: response
  description: >-
    Every API response carries an `x-request-id` header (a UUID). Claid support asks for this value
    when reporting an image processing issue.
  example: 'x-request-id: 4005f581-723f-4c79-85a1-bdd752718887'
  support_contact: support@claid.ai
rate_limit_signaling:
  headers:
  - RateLimit-Limit
  - RateLimit-Remaining
  - RateLimit-Reset
  style: >-
    RateLimit-Limit is a list of policies, e.g. `120, 120;w=60, 4;w=1`, where the first value is the
    limit currently in effect and each `n;w=s` pair is n requests per s seconds.
  exceeded_status: 429
  artifact: rate-limits/lets-enhance-rate-limits.yml
  docs: https://docs.claid.ai/rate-limits
errors:
  format: proprietary
  envelope_fields:
  - error_code
  - error_type
  - error_message
  - error_details
  problem_json: false
  artifact: errors/lets-enhance-error-codes.yml
  docs: https://docs.claid.ai/errors
webhooks:
  supported: true
  artifact: asyncapi/lets-enhance-claid-webhooks.yml
  signature_header: X-Claid-Hmac-SHA256
  signature_algorithm: HMAC-SHA256
  configuration_url: https://claid.ai/account/webhook
storage:
  model: >-
    Results are returned as temporary URLs by default, or written to a storage connector the account
    has registered. Supported connector types are AWS S3, Google Cloud Storage and web folders.
  docs: https://docs.claid.ai/storage-connectors/overview
related:
  authentication: authentication/lets-enhance-authentication.yml
  scopes: scopes/lets-enhance-scopes.yml
  errors: errors/lets-enhance-error-codes.yml
  rate_limits: rate-limits/lets-enhance-rate-limits.yml
  lifecycle: lifecycle/lets-enhance-lifecycle.yml