LINDAT UDPipe API

# Provenance: DERIVED by API Evangelist from the operator's own published REST API
# reference, then verified against live responses from the production host.
# UFAL publishes human documentation, not a machine-readable contract, for this service.
# method: derived
# x-source-url: https://lindat.mff.cuni.cz/services/udpipe/api-reference.php
# generated: 2026-09-01
# x-operator: institution (Institute of Formal and Applied Linguistics, Faculty of
#   Mathematics and Physics, Charles University — host lindat.mff.cuni.cz is the
#   institution's own registrable domain)
openapi: 3.2.0
info:
  title: LINDAT UDPipe API
  version: '2'
  summary: Tokenization, POS tagging, lemmatization and dependency parsing operated by ÚFAL at Charles University.
  description: >-
    Public REST API for UDPipe, a trainable pipeline for tokenization, part-of-speech tagging,
    lemmatization and dependency parsing of CoNLL-U files, operated by the Institute of Formal
    and Applied Linguistics (ÚFAL), Faculty of Mathematics and Physics, Charles University, and
    delivered through the LINDAT/CLARIAH-CZ research infrastructure. The live service advertised
    961 Universal Dependencies models when probed. No API key is required.
    Derived by API Evangelist from the published REST API reference at
    https://lindat.mff.cuni.cz/services/udpipe/api-reference.php and verified against live
    responses from https://lindat.mff.cuni.cz/services/udpipe/api on 2026-09-01.
  contact:
    name: LINDAT/CLARIAH-CZ Helpdesk
    email: lindat-help@ufal.mff.cuni.cz
    url: https://lindat.mff.cuni.cz/user_feedback
  license:
    name: CC BY-NC-SA
    url: https://creativecommons.org/licenses/by-nc-sa/4.0/
  termsOfService: https://lindat.mff.cuni.cz/en/terms-of-use
servers:
  - url: https://lindat.mff.cuni.cz/services/udpipe/api
    description: Production service operated by ÚFAL, Charles University
tags:
  - name: udpipe
    description: Universal Dependencies processing operations
paths:
  /models:
    get:
      operationId: listUDPipeModels
      summary: List available UDPipe models
      description: Returns every model available in the UDPipe REST API with the components it carries (tokenizer, tagger, parser), plus the default model.
      tags: [udpipe]
      responses:
        '200':
          description: Model inventory
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ModelList'
  /process:
    get:
      operationId: processText
      summary: Process supplied data through the UDPipe pipeline
      tags: [udpipe]
      parameters:
        - name: data
          in: query
          required: true
          description: Input text in UTF-8.
          schema:
            type: string
        - name: model
          in: query
          required: false
          description: Model to use. Accepts a full model name, a name without the version suffix, a model-name prefix, or an ISO 639-1 / ISO 639-2 language code such as `cs`, `ces` or `en`.
          schema:
            type: string
        - name: tokenizer
          in: query
          required: false
          description: When present, the input is treated as plain text and tokenized. A value is passed through as tokenizer options.
          schema:
            type: string
        - name: input
          in: query
          required: false
          description: Input format used when the tokenizer is not engaged.
          schema:
            type: string
            enum: [conllu, generic_tokenizer, horizontal, vertical]
            default: conllu
        - name: tagger
          in: query
          required: false
          description: When present, the input is POS tagged and lemmatized. A value is passed through to the tagger.
          schema:
            type: string
        - name: parser
          in: query
          required: false
          description: When present, the input is dependency parsed. A value is passed through to the parser.
          schema:
            type: string
        - name: output
          in: query
          required: false
          description: Output format.
          schema:
            type: string
            enum: [conllu, horizontal, matxin, plaintext, vertical]
            default: conllu
      responses:
        '200':
          description: Processed output
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ServiceResponse'
    post:
      operationId: processTextPost
      summary: Process supplied data (form-encoded or file upload)
      tags: [udpipe]
      requestBody:
        required: true
        content:
          application/x-www-form-urlencoded:
            schema:
              $ref: '#/components/schemas/ProcessRequest'
          multipart/form-data:
            schema:
              $ref: '#/components/schemas/ProcessRequest'
      responses:
        '200':
          description: Processed output
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ServiceResponse'
components:
  schemas:
    ModelList:
      type: object
      properties:
        models:
          type: object
          description: Map of model name to the pipeline components that model provides.
          additionalProperties:
            type: array
            items:
              type: string
              enum: [tokenizer, tagger, parser]
        default_model:
          type: string
      required: [models, default_model]
    ServiceResponse:
      type: object
      description: Common LINDAT web-service response envelope.
      properties:
        model:
          type: string
        acknowledgements:
          type: array
          items:
            type: string
            format: uri
        result:
          type: string
          description: UDPipe output in the requested output format.
      required: [model, result]
    ProcessRequest:
      type: object
      properties:
        data:
          type: string
        model:
          type: string
        tokenizer:
          type: string
        input:
          type: string
          enum: [conllu, generic_tokenizer, horizontal, vertical]
        tagger:
          type: string
        parser:
          type: string
        output:
          type: string
          enum: [conllu, horizontal, matxin, plaintext, vertical]
      required: [data]