Groq Reranking API

openapi: 3.0.1
info:
  title: GroqCloud Audio Reranking API
  description: Specification of the Groq cloud API
  termsOfService: https://groq.com/terms-of-use/
  contact:
    name: Groq Support
    email: support@groq.com
  version: '2.1'
servers:
- url: https://api.groq.com
security:
- api_key: []
tags:
- name: Reranking
paths:
  /openai/v1/reranking:
    post:
      operationId: createReranking
      tags:
      - Reranking
      summary: Reranks documents based on their relevance to a query.
      description: 'Given a query and a list of documents, returns the documents ranked by their relevance to the query.

        The documents are scored and sorted in descending order of relevance.

        '
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/RerankingRequest'
      responses:
        '200':
          description: OK
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RerankingResponse'
      x-groq-metadata:
        returns: "A list of documents sorted by relevance score in descending order. \nScores range from 0.0 to 1.0, where higher scores indicate greater relevance to the query.\n"
        examples:
        - title: Basic Reranking
          request:
            curl: "curl https://api.groq.com/openai/v1/reranking \\\n  -H \"Authorization: Bearer $GROQ_API_KEY\" \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\n    \"model\": \"qwen3-reranker-4b\",\n    \"query\": \"artificial intelligence\",\n    \"docs\": [\n      \"Machine learning is a subset of AI\",\n      \"The weather is nice today\",\n      \"Deep learning uses neural networks\"\n    ]\n  }'\n"
            py: "import os\nfrom groq import Groq\n\nclient = Groq(api_key=os.environ.get(\"GROQ_API_KEY\"))\n\nreranking = client.reranking.create(\n    model=\"qwen3-reranker-4b\",\n    query=\"artificial intelligence\",\n    docs=[\n        \"Machine learning is a subset of AI\",\n        \"The weather is nice today\", \n        \"Deep learning uses neural networks\"\n    ]\n)\nprint(reranking.results)\n"
            js: "import Groq from 'groq-sdk';\n\nconst client = new Groq({\n  apiKey: process.env['GROQ_API_KEY'],\n});\n\nasync function main() {\n  const reranking = await client.reranking.create({\n    model: 'qwen3-reranker-4b',\n    query: 'artificial intelligence',\n    docs: [\n      'Machine learning is a subset of AI',\n      'The weather is nice today',\n      'Deep learning uses neural networks'\n    ]\n  });\n  console.log(reranking.results);\n}\nmain();\n"
          response: "{\n  \"results\": [\n    {\n      \"doc\": \"Machine learning is a subset of AI\",\n      \"score\": 0.92\n    },\n    {\n      \"doc\": \"Deep learning uses neural networks\", \n      \"score\": 0.87\n    },\n    {\n      \"doc\": \"The weather is nice today\",\n      \"score\": 0.23\n    }\n  ]\n}\n"
        - title: Reranking with Custom Instruction
          request:
            json: "{\n  \"model\": \"qwen3-reranker-4b\",\n  \"query\": \"climate change effects\",\n  \"docs\": [\n    \"Global warming causes sea level rise\",\n    \"Electric cars reduce emissions\",\n    \"Renewable energy is growing fast\"\n  ],\n  \"instruction\": \"Find documents specifically about environmental impacts\"\n}\n"
          response: "{\n  \"results\": [\n    {\n      \"doc\": \"Global warming causes sea level rise\",\n      \"score\": 0.95\n    },\n    {\n      \"doc\": \"Electric cars reduce emissions\",\n      \"score\": 0.78\n    },\n    {\n      \"doc\": \"Renewable energy is growing fast\",\n      \"score\": 0.65\n    }\n  ]\n}\n"
components:
  schemas:
    RerankingResult:
      type: object
      properties:
        doc:
          type: string
          description: The original document text.
          example: Machine learning is a subset of artificial intelligence
        score:
          type: number
          format: float
          minimum: 0
          maximum: 1
          description: "Relevance score between 0.0 and 1.0, where higher scores indicate \ngreater relevance to the query.\n"
          example: 0.92
      required:
      - doc
      - score
    RerankingRequest:
      type: object
      additionalProperties: false
      properties:
        model:
          description: 'ID of the reranking model to use.

            '
          example: qwen3-reranker-4b
          type: string
        query:
          description: 'The search query to rank documents against.

            '
          example: artificial intelligence research
          type: string
        docs:
          description: 'An array of documents to rank. Each document is a string containing the text content.

            Maximum of 100 documents per request.

            '
          type: array
          minItems: 1
          maxItems: 100
          items:
            type: string
            minLength: 1
          example:
          - Machine learning is a subset of artificial intelligence
          - The weather forecast predicts rain tomorrow
          - Deep learning uses neural networks with multiple layers
        instruction:
          description: "Optional instruction to guide the reranking process. If not provided, \na default instruction will be used.\n"
          example: Find the most relevant document about AI research
          type: string
          nullable: true
      required:
      - model
      - query
      - docs
    RerankingResponse:
      type: object
      properties:
        results:
          type: array
          description: 'List of documents sorted by relevance score in descending order.

            Each result contains the original document text and its relevance score.

            '
          items:
            $ref: '#/components/schemas/RerankingResult'
      required:
      - results
  securitySchemes:
    api_key:
      type: http
      scheme: bearer
      bearerFormat: apiKey
x-groq-metadata:
  groups:
  - id: chat
    type: endpoints
    title: Chat
    description: ''
    sections:
    - type: endpoint
      key: createChatCompletion
      path: create
  - id: responses
    type: endpoints
    title: Responses (beta)
    description: ''
    sections:
    - type: endpoint
      key: createResponse
      path: create
  - id: audio
    type: endpoints
    title: Audio
    description: ''
    sections:
    - type: endpoint
      key: createTranscription
      path: transcription
    - type: endpoint
      key: createTranslation
      path: translation
    - type: endpoint
      key: createSpeech
      path: speech
  - id: models
    type: endpoints
    title: Models
    description: ''
    sections:
    - type: endpoint
      key: listModels
      path: list
    - type: endpoint
      key: retrieveModel
      path: retrieve
  - id: batches
    type: endpoints
    title: Batches
    description: ''
    sections:
    - type: endpoint
      key: createBatch
      path: create
    - type: endpoint
      key: retrieveBatch
      path: retrieve
    - type: endpoint
      key: listBatches
      path: list
    - type: endpoint
      key: cancelBatch
      path: cancel
  - id: files
    type: endpoints
    title: Files
    description: ''
    sections:
    - type: endpoint
      key: uploadFile
      path: upload
    - type: endpoint
      key: listFiles
      path: list
    - type: endpoint
      key: deleteFile
      path: delete
    - type: endpoint
      key: retrieveFile
      path: retrieve
    - type: endpoint
      key: downloadFile
      path: download
  - id: fine-tuning
    type: endpoints
    title: Fine Tuning
    description: ''
    sections:
    - type: endpoint
      key: listFineTunings
      path: list
    - type: endpoint
      key: createFineTuning
      path: create
    - type: endpoint
      key: getFineTuning
      path: get
    - type: endpoint
      key: deleteFineTuning
      path: delete