--- title: "Update knowledge base" method: PUT path: "/assistant_knowledges/{knowledge_id}" tags: ["Knowledge Bases"] --- # Update knowledge base `PUT /assistant_knowledges/{knowledge_id}` Update a specific knowledge base configuration ## Path parameters - `knowledge_id` string, required ## Request body - KnowledgeBaseUpdateRequest — Request body for updating a knowledge base - `name` string — Name of the knowledge base - `source` KBSourceConfig — Sources configuration - `sources` KBSource[] - `_id` string — Unique identifier for the source - `type` 'web-site' | 'document' | 'plain-text', required — Type of source - `data` string, required — URL for web-site, file URL for document, or text content for plain-text - `name` string — Name of the document (for document type) - `metadata` string — Additional metadata - `url` string — Alternative URL field - `pages` KBWebPage[] — Scraped pages (for web-site type) - `_id` string - `src` string — URL of the page - `title` string — Page title - `scrapping_status` boolean — Whether scraping was successful - `LLM` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `reranking` KBRerankingConfig — Reranking configuration - `weighting_multiplier_coefficient` number, float — Coefficient for weighting multiplier - `chunking` KBChunkingConfig — Chunking configuration for different file types - `web` object — Chunking configuration for web content - `chunking_method` 'fixed_size' | 'dynamic' — Method for chunking web content - `chunk_size` integer — Size of chunks in characters - `chunk_overlap` integer — Overlap between chunks in characters - `instructions` string — Custom instructions for chunking - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `image` object — Chunking configuration for images - `chunking_method` 'fixed_size' | 'dynamic' — Method for chunking image content - `chunk_size` integer - `chunk_overlap` integer - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `pdf` object — Chunking configuration for PDF files - `chunking_method` 'fixed_size' | 'dynamic' | 'by_page' — Method for chunking: fixed_size, dynamic, or by_page (PDF only) - `chunk_size` integer - `chunk_overlap` integer - `split_pdf_pages` boolean — Whether to split PDF by pages - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `docx` object — Chunking configuration for DOCX files - `chunking_method` 'fixed_size' | 'dynamic' | 'by_page' — Method for chunking: fixed_size, dynamic, or by_page (DOCX only) - `chunk_size` integer - `chunk_overlap` integer - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `ppt` object — Chunking configuration for PPT files - `chunking_method` 'fixed_size' | 'dynamic' | 'by_page' — Method for chunking: fixed_size, dynamic, or by_page (PPT only) - `chunk_size` integer - `chunk_overlap` integer - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `txt` object — Chunking configuration for TXT files - `chunking_method` 'fixed_size' | 'dynamic' — Method for chunking text files - `chunk_size` integer - `chunk_overlap` integer - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `csv` object — Chunking configuration for CSV files - `chunking_method` 'by_line' — CSV files only support by_line chunking method - `chunk_size` integer - `chunk_overlap` integer - `instructions` string - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `ingestion` KBIngestionConfig — Ingestion configuration for different file types - `delay_between_documents_ms` integer — Delay between processing documents in milliseconds - `web` object — Web scraping ingestion configuration - `waiting_time_for_web_scraping` integer — Wait time in milliseconds before scraping - `library` 'Cheerio' | 'Puppeteer' — Scraping library to use - `selector` string — CSS selector for content extraction - `exclude` string — CSS selector for elements to exclude - `proxy` object — Proxy configuration for web scraping - `status` boolean — Whether proxy is enabled - `country` string — Country for proxy - `image` object — Image ingestion configuration - `strategy` 'parse' | 'ocr' — Processing strategy for images - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `pdf` object — PDF ingestion configuration - `strategy` 'parse' | 'ocr' — Strategy for PDF processing: parse (native text) or ocr (scanned docs) - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `docx` object — DOCX ingestion configuration - `strategy` 'parse' | 'ocr' — Processing strategy for DOCX files - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `ppt` object — PPT ingestion configuration - `strategy` 'parse' | 'ocr' — Processing strategy for PPT files - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `txt` object — TXT ingestion configuration - `strategy` 'parse' | 'ocr' — Processing strategy for TXT files - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `csv` object — CSV ingestion configuration - `strategy` 'parse' | 'ocr' — Processing strategy for CSV files - `instructions` string — Custom instructions for processing - `llm_model` KBLLMConfig — LLM model configuration for the knowledge base - `model_id` string - `model_name` string - `provider` string - `name` string - `is_document_confidential` boolean — Whether documents are confidential ## Response `200` Knowledge base successfully updated - Success - `success` boolean ## Other responses - `400` — Invalid request - `401` — Unauthorized - Invalid or missing authentication - `404` — Knowledge base not found --- [API](https://skmtc.net/smartly/apis/smartly-ai-apis.md) · [All operations](https://skmtc.net/smartly/apis/smartly-ai-apis/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/smartly/smartly-ai-apis/revisions/1e26909d86c9/schema)