HKUDS/LightRAG

Set these on your host before deploying. Each platform has an “Environment Variables” section in the deploy flow. (showing first 20, more exist in env.example)

• HOST Server Configuration
• PORT
• WEBUI_TITLE
• WEBUI_DESCRIPTION
• OLLAMA_EMULATING_MODEL_TAG Ollama Emulating Model and Tag OLLAMA_EMULATING_MODEL_NAME=lightrag
• ENABLE_LLM_CACHE Query Configuration How to control the context length sent to LLM: MAX_ENTITY_TOKENS + MAX_RELATION_TOKENS < MAX_TOTAL_TOKENS Chunk_Tokens…
• ENABLE_CONTENT_HEADINGS Append each chunk's heading path (parent headings joined by " → ") as a `content_headings` field in the chunk JSON sent to the LLM. Costs…
• RERANK_BINDING Reranking configuration RERANK_BINDING type: null, cohere, jina, aliyun For rerank model deployed by vLLM use cohere binding If LightRAG…
• SUMMARY_LANGUAGE Document processing configuration Document processing output language: English, Chinese, French, German ...
• ENTITY_EXTRACTION_USE_JSON Enable JSON-structured output for entity extraction Default behavior: JSON output is disabled when ENTITY_EXTRACTION_USE_JSON is unset JSON…
• LIGHTRAG_PARSER Multimodal parsing/analyze integration Optional parser routing rules. Example for VLM & MinerU enabled configuration:…
• NATIVE_MD_IMAGE_DOWNLOAD_ENABLED Native Markdown (.md / .textpack) remote image handling External http(s) images in markdown are downloaded and embedded into the sidecar…
• MINERU_API_MODE MinerU shared parameters (both local and official modes) MinerU API protocol. Choose one active mode. official: MinerU precision API v4.…
• MINERU_LOCAL_ENDPOINT MinerU local-only (MINERU_API_MODE=local)
• MINERU_LOCAL_BACKEND MINERU_LOCAL_BACKEND: which mineru-api backend handles the parse. Accepted values (per mineru-api POST /tasks form parameter `backend`):…
• MINERU_LOCAL_PARSE_METHOD MINERU_LOCAL_PARSE_METHOD: parsing strategy for the pipeline component. Accepted values: auto auto-detect embedded text-layer vs OCR per…
• MINERU_LOCAL_IMAGE_ANALYSIS MINERU_LOCAL_IMAGE_ANALYSIS: enable VLM image/chart analysis pass for better caption an footnote recognition. Only consumed by…
• DOCLING_ENDPOINT Docling parser (docling-serve v1 / async API). Endpoint: base URL only — the client appends /v1/convert/file/async,…
• DOCLING_DO_OCR
• DOCLING_FORCE_OCR DOCLING_FORCE_OCR can be overridden per file via a hint / rule on the engine token, e.g. scan.[docling(force_ocr=true)].pdf