> ## Documentation Index > Fetch the complete documentation index at: https://io-migration.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # RAG Query > Execute a RAG (Retrieval-Augmented Generation) query. ## OpenAPI ````yaml openapi/retrieve-and-generate-rag.json post /api/r2r/v3/retrieval/rag openapi: 3.1.0 info: title: IO Intelligence version: '1.0' servers: - url: https://api.intelligence.io.solutions security: - sec0: [] paths: /api/r2r/v3/retrieval/rag: post: summary: RAG Query description: Execute a RAG (Retrieval-Augmented Generation) query. operationId: retrieve-and-generate-rag requestBody: content: application/json: schema: type: object properties: query: type: string description: The user's question search_mode: type: string description: >- Default value of custom allows full control over search settings. Pre-configured search modes: basic: A simple semantic-based search. advanced: A more powerful hybrid search combining semantic and full-text. custom: Full control via search_settings. If filters or limit are provided alongside basic or advanced, they will override the default settings for that mode. default: custom enum: - basic - advanced - custom search_settings: type: object description: >- The search configuration object. If search_mode is custom, these settings are used as-is. For basic or advanced, these settings will override the default mode configuration. Common overrides include filters to narrow results and limit to control how many results are returned. properties: use_hybrid_search: type: boolean description: >- Whether to perform a hybrid search. This is equivalent to setting use_semantic_search=True and use_fulltext_search=True, e.g. combining vector and keyword search. default: false use_semantic_search: type: boolean description: Whether to use semantic search default: true use_fulltext_search: type: boolean description: Whether to use full-text search default: false filters: type: string description: >- Filters to apply to the search. Allowed operators include eq, neq, gt, gte, lt, lte, like, ilike, in, and nin. Commonly seen filters include operations include the following: {"document_id": {"$eq": "9fbe403b-..."}} {"document_id": {"$in": ["9fbe403b-...", "3e157b3a-..."]}} {"collection_ids": {"$overlap": ["122fdf6a-...", "..."]}} {"$and": {"$document_id": ..., "collection_ids": ...}} limit: type: integer description: Maximum number of results to return. >=1 <=1000 default: 10 format: int32 offset: type: string description: Offset to paginate search results default: '0' include_metadatas: type: boolean description: >- Whether to include element metadata in the search results default: true include_scores: type: boolean description: >- Whether to include search score values in the search results default: true search_strategy: type: string description: >- Search strategy to use (e.g., ‘vanilla’, ‘query_fusion’, ‘hyde’) default: vanilla hybrid_settings: type: object description: >- Settings for hybrid search (only used if use_semantic_search and use_fulltext_search are both true) properties: full_text_weight: type: number description: Weight to apply to full text search default: 1 format: double semantic_weight: type: number description: Weight to apply to semantic search default: 5 format: double full_text_limit: type: integer description: >- Maximum number of results to return from full text search default: 200 format: int32 rrf_k: type: integer description: K-value for RRF (Rank Reciprocal Fusion) default: 50 format: int32 chunk_settings: type: object description: Settings specific to chunk/vector search properties: index_measure: type: string description: The distance measure to use for indexing enum: - l2_distance - max_inner_product - cosine_distance - l1_distance - hamming_distance - jaccard_distance probes: type: integer description: >- Number of ivfflat index lists to query. Higher increases accuracy but decreases speed. default: 10 format: int32 ef_search: type: integer description: >- Size of the dynamic candidate list for HNSW index search. Higher increases accuracy but decreases speed. default: 40 format: int32 enabled: type: boolean description: Whether to enable chunk search default: true graph_settings: type: object description: Settings specific to knowledge graph search properties: limits: type: array enabled: type: boolean description: Whether to enable graph search default: true num_sub_queries: type: integer description: >- Number of sub-queries/hypothetical docs to generate when using hyde or rag_fusion search strategies. default: 5 format: int32 rag_generation_config: type: object description: Configuration for RAG generation properties: model: type: string temperature: type: number format: double top_p: type: number format: double max_tokens_to_sample: type: integer format: int32 stream: type: boolean functions: type: array tools: type: array add_generation_kwargs: type: array api_base: type: string response_format: type: array items: properties: Base Model: type: object type: object extended_thinking: type: boolean description: >- Flag to enable extended thinking mode (for Anthropic providers) default: false thinking_budget: type: integer description: >- Token budget for internal reasoning when extended thinking mode is enabled. Must be less than max_tokens_to_sample. format: int32 reasoning_effort: type: string description: >- Effort level for internal reasoning when extended thinking mode is enabled, low, medium, or high.Only applicable to OpenAI providers. task_prompt: type: string description: Optional custom prompt to override default include_title_if_available: type: boolean description: Include document titles in responses when available default: false include_web_search: type: boolean description: Include web search results provided to the LLM. default: false responses: '200': description: '200' content: application/json: examples: Result: value: key: value schema: type: object properties: key: type: string example: value '404': description: '404' content: application/json: examples: Result: value: {} schema: type: object properties: {} '422': description: '422' content: text/plain: examples: Result: value: '' deprecated: false components: securitySchemes: sec0: type: oauth2 flows: {} ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.