v1.5.0

## Major Improvements

*  🎉✨**Feacture**:  **LightRAG now supports multimodal document processing** and can fully leverage images, tables, and formulas within documents to answer queries. **All RagAnything’s multimodal processing capabilities are merged into LightRAG**; ( [RagAnything](https://github.com/HKUDS/RAG-Anything) will no longer receive core feature updates or maintenance going forward)
*  ✨🎉**Feacture**: **Fully upgraded the document processing pipeline**, with support for using **MinerU** and **Docling** to parse and extract file content, seamlessly integrating with multimodal content analysis and entity-relationship extraction.
*  💡✨**Feacture**: LightRAG now introduces **Native Parsing**, enabling high-quality content extraction from DOCX documents. It supports accurate reconstruction of Word auto-numbering, as well as extraction of images, tables, and formulas, providing seamless integration with multimodal content analysis and entity-relationship extraction. Expanded format support for the **Native Parser** is coming soon.
*  💡🧠**Feacture**: **Introduced four selectable text chunking strategies: `Fix`, `Recursive`, `Vector`, and `Paragraph`.** The parameters for each chunking strategy can be configured through environment variables.
*  💡🎯**Feacture**: The file processing pipeline supports **selecting the file parsing engine and text chunking strategy either based on file extensions or on a per-file basis**. For detailed usage instructions, refer to [FileProcessingPipeline.md](https://github.com/HKUDS/LightRAG/blob/main/docs/FileProcessingPipeline.md).
*  🚀⚡**Performance**: Optimized the vector storage persistence logic by deferring vector computations until the end of each file processing batch, enabling centralized bulk computation. **This significantly reduced the number of vector model invocations and significantly enhance the upsert speed of all vector DB LightRAG supported**.
*  Enable **task-aware embedding** support for asymmetric models, including `voyage-3`, `text-embedding-004`, `embed-multilingual-v3.0`, and `jina-embeddings-v3`.
*  Improved entity/relation extraction reliability by **introducing LLM  JSON-structured output**; set ENTITY_EXTRACTION_USE_JSON=true to enable it.
*  Introduce `ENTITY_TYPE_PROMPT_FILE` to empower users with enhanced guidance for LLM-driven **entity type recognition** and extraction.
*  Fully support Amazon and Anthropic models on **AWS Bedrock API**.
*  🎯✨🎉**Feacture**: Implemented **role-specific LLM configuration** support, introducing four distinct roles: EXTRACT, QUERY, KEYWORDS, and VLM, each with independent LLM settings. It is recommended to configure the KEYWORDS role with a small-parameter, non-reasoning, high-speed model to optimize query latency; the EXTRACT role with a medium-parameter, non-reasoning model to balance accuracy and throughput; and the QUERY role with a large-parameter reasoning model to enhance query quality. For detailed usage instructions, refer to [RoleSpecificLLMConfiguration.md](https://github.com/HKUDS/LightRAG/blob/main/docs/RoleSpecificLLMConfiguration.md).

## What's Broken

* 💥Warning: Upgrades require careful planning for users currently utilizing OpenSearch and MongoDB graph storage, **as data migration and system restarts may cause extended service outages**.
* ⚠️**The `ENTITY_TYPES` environment variable has been deprecated**; Please replace it with ENTITY_TYPE_PROMPT_FILE before launching this new version.
* 💔**Removed deprecated field**: Deleted `QueryParam.model_func` from `lightrag/base.py`
* For OpenSearch versions prior to 3.3.0, upgrade OpenSearch before upgrading LightRAG (see https://github.com/HKUDS/LightRAG/pull/2991).
* ⚠️Move delete_entity/delete_relation from document API to graph API

| Before                              | After                           |
| ----------------------------------- | ------------------------------- |
| `DELETE /documents/delete_entity`   | `DELETE /graph/entity/delete`   |
| `DELETE /documents/delete_relation` | `DELETE /graph/relation/delete` |

## What's Changed

* feat: integrate structured extraction and multimodal role-based pipelineFeat/multimodal pipeline by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2830
* ♻️ refactor(documentManager): reorganize document status filtering by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2851
* Remove config.ini from compose samples by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2906
* fix: handle OpenAI length finish reason fallback by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2913
* feat: apply entity extraction best practice and add full-service comparisons by @MrGidea in https://github.com/HKUDS/LightRAG/pull/2914
* Feature/enhance entity extraction stability dev by @yunzhongxiaxi in https://github.com/HKUDS/LightRAG/pull/2864
* feat(extraction): configurable per-response entity/relation limits by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2950
* ♻️ refactor(llm): unify keyword extraction across providers by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2953
* ♻️ refactor(llm): unify structured output control via response_format by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2956
* refactor(gemini): improve default endpoint handling and sdk integration by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2957
* refactor(bedrock): support default and custom endpoints by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2958
* refactor(setup): use sentinel endpoints for Gemini and Bedrock defaults by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2959
* perf(postgres): use binary parameter for vector similarity queries by @wkpark in https://github.com/HKUDS/LightRAG/pull/2949
* feat(prompt): externalize entity type extraction profiles by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2964
* feat(bedrock): rename aws_bedrock to bedrock and add BindingOptions support by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2966
* chore(deps): bump react-router-dom from 7.14.0 to 7.14.1 in /lightrag_webui in the react group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/2967
* chore(deps-dev): bump the build-tools group in /lightrag_webui with 3 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/2968
* chore(deps): bump the frontend-minor-patch group across 1 directory with 3 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/2969
* Fix role LLM max async fallback by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2973
* fix(llm): tighten client and stream cleanup across LLM bindings by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2974
* chore(deps): bump lucide-react from 0.577.0 to 1.6.0 in /lightrag_webui by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/2970
* docs: add role-specific LLM configuration guide by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2976
* refactor: unify role LLM config via ROLES registry + queue observability  by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2978
* feat(status): role-based LLM observability and storage workspace info by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2980
* feat(rerank): add independent concurrency and timeout configuration by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2981
* Fix LLM cache role identity isolation by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2982
* Add role LLM provider options logging and change role provider options to start from empty not default by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2984
* Add Podman-compatible compose file by @tears710 in https://github.com/HKUDS/LightRAG/pull/2983
* Fix bedrock/gemini host leak from env.example on make server/storage by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2985
* fix: remove `stream` parameter from `.parse()` call when `response_format` is present by @PaulTitto in https://github.com/HKUDS/LightRAG/pull/2965
* Explicit voyageai embed support by @laszukdawid in https://github.com/HKUDS/LightRAG/pull/2484
* feat: Add task-aware embedding support by @StoreksFeed in https://github.com/HKUDS/LightRAG/pull/2560
* Improve DOCX parsing idempotency by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2987
* OpenSearch: Use version-aware sort tiebreaker for PIT search by @LantaoJin in https://github.com/HKUDS/LightRAG/pull/2991
* Refactor DOCX archive handling by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2994
* refactor: remove embedded docling and raganything fallbacks by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2997
* Refactor file parser routing by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2998
* Deduplicate documents by filename by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3000
* Fix parsed document artifact isolation by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3005
* Fix parser-hinted filename dedup by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3006
* feat(native): align docx parser with mineru LIGHTRAG output (.blocks.jsonl) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3008
* ♻️ refactor(lightrag): split monolithic lightrag.py into focused modules by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3012
* feat(parser_routing): per-file process options and canonical basename by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3013
* feat(api): pipeline reentrancy guards and idempotent multimodal analyze by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3014
* fix(pipeline): preserve process_options in doc_status metadata across transitions by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3017
* feat(pipeline): resume already-extracted documents under current process_options by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3015
* refactor(pipeline): normalize parser / extraction metadata field names by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3021
* ♻️ refactor(pipeline): unify F chunking for raw and lightrag formats by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3023
* ♻️ refactor(docx): move native parser to lightrag/native_parser/docx by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3030
* chore(deps): bump react-router-dom from 7.14.1 to 7.14.2 in /lightrag_webui in the react group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3001
* chore(deps): bump the ui-components group in /lightrag_webui with 3 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3002
* chore(deps): bump axios from 1.15.1 to 1.15.2 in /lightrag_webui in the frontend-minor-patch group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3004
* chore(deps-dev): bump the build-tools group across 1 directory with 5 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3003
* Replace status tooltip with details modal by @g2424303264-code in https://github.com/HKUDS/LightRAG/pull/3025
* fix(webui): resolve react-hooks lint errors after eslint-plugin upgrade by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3036
* chore(deps): bump lucide-react from 1.9.0 to 1.14.0 in /lightrag_webui in the ui-components group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3032
* chore(deps): bump sigma from 3.0.2 to 3.0.3 in /lightrag_webui in the graph-viz group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3033
* chore(deps-dev): bump typescript-eslint from 8.59.1 to 8.59.2 in /lightrag_webui in the build-tools group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3034
* chore(deps): bump i18next from 25.10.10 to 26.0.3 in /lightrag_webui by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3035
* Add prefix path to API and WebUI by @shahrin014 in https://github.com/HKUDS/LightRAG/pull/3007
* feat(webui): inject path prefix at runtime — one build serves all sites by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3039
* refactor(pipeline): split apipeline_process_enqueue_documents into helpers by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3041
* feat(openai): inject X-DashScope-Workspace header from DASHSCOPE_WORKSPACE_ID env by @getsov75-maker in https://github.com/HKUDS/LightRAG/pull/3042
* Update README by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3047
* Add paragraph semantic chunking strategy by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3044
* feat(chunker): add R/V chunkers and chunk_options snapshot mechanism by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3046
* Add offline sample retrieval check for RAGAS evaluation by @FU-max-boop in https://github.com/HKUDS/LightRAG/pull/3038
* fix(chunker): CJK punctuation support and chunk_token_size enforcement by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3050
* feat(chunker): split oversized tables on row boundaries before char fallback by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3051
* chore(deps): bump sigstore/cosign-installer from 4.1.1 to 4.1.2 in the github-actions group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3053
* fix(chunker): preserve HTML table row group wrappers by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3055
* fix(PGGraphStorage): edge properties lost on upsert with Apache AGE by @oleksandr-kushnir in https://github.com/HKUDS/LightRAG/pull/3052
* feat(multimodal): backfill surrounding context on native sidecars by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3057
* docs: fix typo 'Papper' to 'Paper' in README by @viraj1995 in https://github.com/HKUDS/LightRAG/pull/3058
* feat(llm/vlm): unify image_inputs across bindings + VLM cache + master switch by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3063
* chore(deps): update google-genai requirement from <2.0.0,>=1.0.0 to >=1.0.0,<3.0.0 by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3062
* chore(deps): update pymilvus requirement from <3.0.0,>=2.6.2 to >=2.6.2,<4.0.0 by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3061
* Refactor multimodal: status semantics, nested chunk schema, per-chunk entity injection by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3064
* style: update EmptyCard and DocumentManager for improved layout by @JackLuguibin in https://github.com/HKUDS/LightRAG/pull/3060
* fix(opensearch): escape wildcard metacharacters in search_labels to prevent DoS (CWE-89) by @sebastiondev in https://github.com/HKUDS/LightRAG/pull/3026
* feat(docx): doc_title from first heading, table_header in sidecar by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3065
* feat(multimodal): defer sidecar surrounding to analyze entry; env-tunable budgets by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3066
* ✨ feat(sidecar): shorten item IDs by stripping doc- prefix by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3067
* feat(multimodal): strip parser-internal markup from sidecar surrounding by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3068
* ✨ feat(extract): enforce MAX_EXTRACT_INPUT_TOKENS for analyze & gleaning by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3073
* chore(deps): bump the react group in /lightrag_webui with 3 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3069
* chore(deps-dev): bump the build-tools group in /lightrag_webui with 3 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3070
* chore(deps): bump the frontend-minor-patch group across 1 directory with 5 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3071
* chore(deps): bump react-i18next from 16.6.6 to 17.0.3 in /lightrag_webui by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3072
* refactor(multimodal): inline bracket-label format for mm chunks by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3074
* fix: extract Docling async markdown result by @he-yufeng in https://github.com/HKUDS/LightRAG/pull/3031
* feat(parse_mineru): unified sidecar writer + MinerU raw bundle cache by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3075
* feat(native_parser/docx): route through unified SidecarWriter by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3077
* feat: dedupe cross-filename uploads via merged_text normalization by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3078
* feat(mineru): split-by-heading block merging with markdown titles by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3079
* test: harden hermetic env fixture and fix MinerU put() fake by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3080
* feat(mineru): emit page-level positions from page_idx by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3081
* feat(docling): route parse_docling through sidecar bundle pipeline by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3085
* fix: guard against IndexError on empty LLM choices list by @qizwiz in https://github.com/HKUDS/LightRAG/pull/3086
* refactor(parser): rename adapters to ir_builder and consolidate parser packages by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3087
* feat(parser_cli): unified debug CLI for native / mineru / docling by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3088
* Tokenizer.encode: gracefully handle disallowed special tokens in content by @RooseveltAdvisors in https://github.com/HKUDS/LightRAG/pull/3082
* Improve MinerU parsing provenance and upload handling by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3089
* fix(parsers): drop empty-bodied tables to prevent analyze worker hard-failure by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3090
* feat(doc-status): add per-stage start time and parse-skipped flag by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3091
* feat(mineru): invalidate raw bundle cache on parser option changes by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3092
* refactor(full-docs): unify path handling and slim chunk_options snapshot by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3093
* refactor(pg): align PG storage fields with JSON storage for parity by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3094
* feat(redis): add basename and content_hash lookups for doc status by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3098
* refactor(mongo): align doc-status storage with JSON storage for parity by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3099
* feat(opensearch): add basename and content_hash lookups for doc status by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3100
* feat(pipeline-status): probe + throttled refresh for prompt scan/upload feedback by @danielaskdd https://github.com/HKUDS/LightRAG/pull/3101
* feat(chunker): give P strategy a dedicated default chunk_token_size by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3102
* Setup wizard: role-specific LLM models, Bedrock auth docs, env tidy by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3116
* ♻️ refactor(parser): consolidate parser modules under lightrag/parser/ by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3117
* build(deps): bump react-router-dom from 7.15.0 to 7.15.1 in /lightrag_webui in the react group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3109
* build(deps): bump the ui-components group in /lightrag_webui with 4 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3110
* feat(pipeline): surface parse/analyze progress to pipeline_status by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3118
* feat(docx-parser): drop revision/comment markers and skip empty tables by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3119
* fix: make EMBEDDING_TOKEN_LIMIT actually enforce token truncation by @kennedydqz-del in https://github.com/HKUDS/LightRAG/pull/3105
* Revert "fix: make EMBEDDING_TOKEN_LIMIT actually enforce token truncation" by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3120
* fix: resolve 3 bugs in Anthropic/Claude LLM provider by @kennedydqz-del in https://github.com/HKUDS/LightRAG/pull/3106
* fix(pipeline): emit per-item analyze logs in real time by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3122
* feat(pipeline): propagate /cancel_pipeline into PARSE and ANALYZE by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3124
* feat(webui): surface parser routing + MinerU/Docling config in status card by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3125
* fix: sync API docs dark theme by @he-yufeng in https://github.com/HKUDS/LightRAG/pull/3123
* write_nx_graph: atomic write via per-writer .tmp + os.replace by @RooseveltAdvisors in https://github.com/HKUDS/LightRAG/pull/3083
* fix(PGGraphStorage): serialise concurrent upsert_edge with advisory lock by @oleksandr-kushnir in https://github.com/HKUDS/LightRAG/pull/3056
* fix(server): normalize scope.path so Mount works in proxy-strip mode by @Xaverrrrr in https://github.com/HKUDS/LightRAG/pull/3128
* fix(concurrency): wrap ainsert_custom_kg writes with keyed lock (follow-up to #3056) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3129
* Remove quotes from status details by @g2424303264-code in https://github.com/HKUDS/LightRAG/pull/3127
* perf(opensearch): batch vector upserts to reduce HTTP roundtrips by @Saswatsusmoy in https://github.com/HKUDS/LightRAG/pull/3043
* perf(opensearch): batch KV operations under namespace lock by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3133
* fix(storage): atomic index_done_callback for Faiss/Json/Nano by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3134
* ♻️ refactor(tests): mirror feature folders under tests/ by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3135
* 🐛 fix(setup): treat opensearch-dashboards as opensearch sub-service by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3136
* Remove deprecated QueryParam.model_func and related code by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3137
* ✨ feat(pipeline): record parse/analyze stage end-time and skipped metadata by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3138
* feat(graph): block graph mutations while pipeline is busy by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3139
* Add PostgreSQL 18 AGE pgvector image by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3140
* feat(setup): allow custom PostgreSQL credentials for Docker deployments by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3141
* feat(api): add chunking strategy selection to /documents/text(s) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3142
* perf(opensearch): defer vector embedding until flush by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3143
* perf(nano): defer vector embedding to flush time by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3145
* perf(faiss): defer vector embedding to flush time by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3147
* 🐛 fix(operate): correct final_data key path in debug log by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3148
* perf(milvus): defer vector embedding until flush by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3149
* perf(mongo): defer vector embedding until flush by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3150
* perf(qdrant): defer vector embedding until flush by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3151
* refact(postgres): defer PGVector embedding to index_done_callback by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3152
* 🐛 fix(kg): post-prune pending buffer in vector delete_entity_relation by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3154
* build(deps): bump the content-rendering group across 1 directory with 2 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3112
* build(deps-dev): bump @types/react from 19.2.14 to 19.2.15 in /lightrag_webui in the react group by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3155
* fix(llm): retry transient OpenAI 5xx and 'could not parse JSON body' 400s by @cvidaillac in https://github.com/HKUDS/LightRAG/pull/3144
* build(deps-dev): bump the build-tools group across 1 directory with 5 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3156
* build(deps): bump the frontend-minor-patch group across 1 directory with 5 updates by @dependabot[bot] in https://github.com/HKUDS/LightRAG/pull/3157
* refactor(parser): make every recognized heading its own block by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3159
* feat(parser): render heading lines with a markdown # prefix by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3160
* 🔧 chore(neo4j): drop @final from Neo4JStorage to allow subclassing by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3163
* fix(sidecar): use source spans for chunk backfill by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3162
* 🐛 fix(milvus): batch flush upserts/deletes under gRPC message limit by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3164
* 🐛 fix(qdrant): resilient flush — own-batch oversized points, batch deletes by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3165
* 🔧 fix(milvus): shrink upsert batch count to 128 (align with Qdrant) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3166
* feat(mongo): batch bulk_write upserts/deletes across all MongoDB storage paths by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3167
* ✨ feat(opensearch): bound bulk write payload/record size by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3168
* ✨ feat(postgres): bound non-graph bulk upsert/delete payload/record size by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3169
* ⚡ feat(postgres): chunk-level transaction fallback for PGGraph batch writes by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3170
* 🔒 fix(opensearch): canonical edge ids to close reciprocal upsert race by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3171
* ⚡️ perf(mongo): exact single-edge reads use canonical (edge_lo, edge_hi) index by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3174
* 🔒 fix(mongo): canonical edge_key + unique index to close upsert duplicate race by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3172
* 🔒 fix(opensearch): merge reverse payload into canonical on duplicate migration by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3173
* ⚡ perf(opensearch): collapse edge point-lookups onto canonical _id by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3175
* 🐛 fix(postgres): rank get_knowledge_graph("*") nodes by undirected degree by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3176
* 🔒 fix(mongo): rebuild FAILED Atlas vector index instead of wedging queries by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3177
* ♻️ refactor(api): move delete_entity/delete_relation to graph API (breaking) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3178
* ♻️ refactor(pipeline): rename metadata fields & add source_file back-compat by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3179
* 🔧 refactor(constants): centralize LLM call priorities and align analysis stage with processing by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3180
* Slim queue payloads by re-reading document bodies from storage by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3182
* chore(config): bump default MAX_PARALLEL_INSERT to 3 by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3183
* Replace sys.exit with exceptions in DOCX parser validation by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3184
* ♻️ refactor(webui): split document status filter tabs by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3185
* Split queue size config into parse and analyze stages by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3186
* 🐛 fix(pipeline): byte-aware entity truncation & fail-fast on storage flush errors by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3187
* 🐛 fix(pipeline): clean stale per-attempt metadata earlier for PENDING docs by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3188
* 🐛 fix(parser): split over-long outline headings at soft line break by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3190
🐛 fix(pipeline): guarantee worker teardown + isolate enqueue get_by_id on backend outage by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3191

## New Contributors

* @yunzhongxiaxi made their first contribution in https://github.com/HKUDS/LightRAG/pull/2864
* @tears710 made their first contribution in https://github.com/HKUDS/LightRAG/pull/2983
* @PaulTitto made their first contribution in https://github.com/HKUDS/LightRAG/pull/2965
* @laszukdawid made their first contribution in https://github.com/HKUDS/LightRAG/pull/2484
* @g2424303264-code made their first contribution in https://github.com/HKUDS/LightRAG/pull/3025
* @shahrin014 made their first contribution in https://github.com/HKUDS/LightRAG/pull/3007
* @getsov75-maker made their first contribution in https://github.com/HKUDS/LightRAG/pull/3042
* @FU-max-boop made their first contribution in https://github.com/HKUDS/LightRAG/pull/3038
* @oleksandr-kushnir made their first contribution in https://github.com/HKUDS/LightRAG/pull/3052
* @viraj1995 made their first contribution in https://github.com/HKUDS/LightRAG/pull/3058
* @JackLuguibin made their first contribution in https://github.com/HKUDS/LightRAG/pull/3060
* @qizwiz made their first contribution in https://github.com/HKUDS/LightRAG/pull/3086
* @kennedydqz-del made their first contribution in https://github.com/HKUDS/LightRAG/pull/3105
* @Xaverrrrr made their first contribution in https://github.com/HKUDS/LightRAG/pull/3128
* @Saswatsusmoy made their first contribution in https://github.com/HKUDS/LightRAG/pull/3043
* @cvidaillac made their first contribution in https://github.com/HKUDS/LightRAG/pull/3144

**Full Changelog**: https://github.com/HKUDS/LightRAG/compare/v1.4.16...v1.5.0