> ## Documentation Index > Fetch the complete documentation index at: https://docs.captide.ai/llms.txt > Use this file to discover all available pages before exploring further. # Chunk Retrieval > Search for relevant document chunks based on a query. Under the hood, Captide's RAG system uses a combination of LLMs and vector databases to find the most relevant chunks of text from corporate disclosures that are most relevant to the query. This endpoint returns chunks of text from corporate disclosures that are most relevant to the query, along with metadata about their source documents. This endpoint is useful for building search interfaces or retrieving specific content from documents. This v2 endpoint supports newer document categories in source mappings (e.g., earnings-presentation, investor-event-presentation, shareholder-meeting-circular, etc.). Agent response quality is exactly the same compared to v1. ## OpenAPI ````yaml GET /api/v2/rag/chunks openapi: 3.1.0 info: title: Captide REST API description: >- API for accessing financial disclosures and AI-powered financial document analysis. version: 0.3.13 servers: - url: https://rest-api.captide.co description: Prod server security: [] paths: /api/v2/rag/chunks: get: tags: - Retrieval Augmented Generation summary: Search document chunks (v2) description: >- Search for relevant document chunks based on a query. Under the hood, Captide's RAG system uses a combination of LLMs and vector databases to find the most relevant chunks of text from corporate disclosures that are most relevant to the query. This endpoint returns chunks of text from corporate disclosures that are most relevant to the query, along with metadata about their source documents. This endpoint is useful for building search interfaces or retrieving specific content from documents. This v2 endpoint supports newer document categories in source mappings (e.g., earnings-presentation, investor-event-presentation, shareholder-meeting-circular, etc.). Agent response quality is exactly the same compared to v1. operationId: search_chunks_api_v2_rag_chunks_get parameters: - name: query in: query required: true schema: type: string minLength: 1 description: The natural language query examples: - What is Zscaler's net dollar retention rate? title: Query description: The natural language query - name: document_ids in: query required: false schema: anyOf: - type: string - type: 'null' description: >- Comma-separated string of document file IDs. If provided and non-empty, skips document and company selection and goes directly to retrieval. title: Document Ids description: >- Comma-separated string of document file IDs. If provided and non-empty, skips document and company selection and goes directly to retrieval. responses: '200': description: Chunks found successfully content: application/json: schema: $ref: '#/components/schemas/ChunksResponse' example: relevantChunks: - pageNumber: 2 documentMetadata: id: 2dfd0c2a-fb6d-486a-ad62-3b1830eb30bf title: ASML 2024 Annual Report documentCategory: annual-financial-report date: '2024-12-31' fiscalQuarter: 4 fiscalYear: 2024 companyName: ASML HOLDING NV tickers: - ASML - ASML.AS '400': description: Invalid request parameters content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '401': description: Invalid or missing API key content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '404': description: Resource not found content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '422': description: Validation error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '429': description: Rate limit exceeded content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '500': description: Internal server error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' security: - APIKeyHeader: [] components: schemas: ChunksResponse: properties: relevantChunks: items: $ref: '#/components/schemas/Chunk' type: array title: Relevantchunks description: List of relevant document chunks type: object required: - relevantChunks title: ChunksResponse description: Response model for chunk search results. ErrorResponse: properties: detail: type: string title: Detail type: object required: - detail title: ErrorResponse description: Standard error response model for all API endpoints. Chunk: properties: content: type: string title: Content description: Text content of the chunk pageNumber: type: integer title: Pagenumber description: Page number in the document excerpt: anyOf: - type: string - type: 'null' title: Excerpt description: >- Specific part of the chunk that the agent used to answer the user's question. Used for source linking and highlighting. documentMetadata: $ref: '#/components/schemas/SourceDocument' description: Metadata for the chunk's document type: object required: - content - pageNumber - documentMetadata title: Chunk description: A chunk of document content with metadata. SourceDocument: properties: id: type: string title: Id description: Unique document ID title: type: string title: Title description: Document title documentCategory: type: string title: Documentcategory description: >- Document categories. For detailed category descriptions, please refer to [this page](/coverage) formType: anyOf: - type: string - type: 'null' title: Formtype description: >- Document SEC form type (only applicable if the document is an SEC filing) date: type: string title: Date description: >- Date the document relates to (e.g., the press release date for a current report, or the fiscal period end date for annual and interim reports) fiscalQuarter: anyOf: - type: integer - type: 'null' title: Fiscalquarter description: Fiscal quarter the document relates to (from 1 to 4) fiscalYear: anyOf: - type: integer - type: 'null' title: Fiscalyear description: Fiscal year the document relates to (e.g., 2024) companyName: type: string title: Companyname description: Name of the company filing the document tickers: items: type: string type: array title: Tickers description: Tickers of the company filing the document metadata: anyOf: - additionalProperties: true type: object - type: 'null' title: Metadata description: Metadata for the document type: object required: - id - title - documentCategory - date - companyName - tickers title: SourceDocument description: Source document metadata model. securitySchemes: APIKeyHeader: type: apiKey in: header name: X-API-Key ````