What's Breaking (for LightRAG Core integration only)

• Rename params of chunking function: If you incorporate the chunking function into LightRAG and pass parameters by name, corresponding code updates are required.

def chunking_by_token_size(
    tokenizer: Tokenizer,
    content: str,
    split_by_character: str | None = None,
    split_by_character_only: bool = False,
    chunk_overlap_token_size: int = 100,
    chunk_token_size: int = 1200,
) -> list[dict[str, Any]]:

• Inject an embedding_func with model_name to LightRAG using wrap_embedding_func_with_attrs:

@wrap_embedding_func_with_attrs(
    embedding_dim=1536, max_token_size=8192, model_name="text-embedding-3-small"
)
@retry(
    stop=stop_after_attempt(3),
    wait=wait_exponential(multiplier=1, min=4, max=60),
    retry=(
        retry_if_exception_type(RateLimitError)
        | retry_if_exception_type(APIConnectionError)
        | retry_if_exception_type(APITimeoutError)
    ),
)
async def embedding_func(texts: list[str]) -> np.ndarray:
    client = AzureOpenAI(
        api_key=AZURE_OPENAI_API_KEY,
        api_version=AZURE_EMBEDDING_API_VERSION,
        azure_endpoint=AZURE_OPENAI_ENDPOINT,
    )
    embedding = client.embeddings.create(model=AZURE_EMBEDDING_DEPLOYMENT, input=texts)

    embeddings = [item.embedding for item in embedding.data]
    return np.array(embeddings)

rag = LightRAG(
      working_dir=WORKING_DIR,
      llm_model_func=llm_model_func,
      embedding_func=embedding_func,
)

To ensure seamless transition, legacy code injecting embedding_func without model_name will continue to interface with the original non-suffixed vector tables.

What's New

• Feat: Add Chain of Thought Support for Gemini LLM by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2326

• Feat: Add Optional Embedding Dimension Control with OpenAI API by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2328

• Feat: Add Gemini Embedding Support to LightRAG by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2329

• Feat: Add LLM Cache Migration Tool by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2330

• Feat: Add LLM Query Cache Cleanup Tool by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2335

• Support async chunking func to improve processing performance when a heavy chunking_func is passed in by user by @tongda in https://github.com/HKUDS/LightRAG/pull/2336

• Add ollama cloud support by @LacombeLouis in https://github.com/HKUDS/LightRAG/pull/2348

• Feat: Add Workspace Isolation for Pipeline Status and In-memory Storage by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2369

• feat: add vchordrq vector index support for PostgreSQL by @wmsnp in https://github.com/HKUDS/LightRAG/pull/2378

• Feat: Enhanced DOCX Extraction with Table Content Support by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2383

• Feat: Enhance XLSX Extraction by Adding Separators and Escape Special Characters by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2386

• Optimize for OpenAI Prompt Caching: Restructure entity extraction pro… by @Ghazi-raad in https://github.com/HKUDS/LightRAG/pull/2426

• feat: Vector Storage Model Isolation with Automatic Migration by @BukeLy in https://github.com/HKUDS/LightRAG/pull/2391

• feat: Implement Vector Database Model Isolation and Auto-Migration by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2513

• feat: Add Automatic Text Truncation Support for Embedding Functions by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2523

What's Changed

• Fix: Remove Duplicate Entity/Realtion Tracking Deletion in adelete_by_doc_id by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2322

• Fix spelling errors in the "使用PostgreSQL存储" section of README-zh.md by @huangbhan in https://github.com/HKUDS/LightRAG/pull/2327

• Add dimensions parameter support to openai_embed() by @yrangana in https://github.com/HKUDS/LightRAG/pull/2323

• Fix Gemini driver retry mechanism by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2331

• HotFix: Restore OpenAI Streaming Response & Refactor keyword_extraction Parameter by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2334

• Refactor: Migrate PDF processing dependency from pypdf2 to actively pypdf by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2338

• Fix: Prevent UnicodeEncodeError in JSON storage operations by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2344

• Remove deprecated response_type parameter from query settings UI by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2345

• Refactor: Optimize write_json for Memory Efficiency and Performance by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2346

• Refact: Remove blocking dependency installation from document upload handlers by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2350

• Refact: Implement Lazy Configuration Initialization for API Server by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2351

• Refact: Enhance DOCLING integration with lazy loading and macOS safeguards by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2352

• Fix: Robust error handling for async database operations in graph storage by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2356

• Update the value corresponding to the extracted entity relationship keywords by @sleeepyin in https://github.com/HKUDS/LightRAG/pull/2358

• Add macOS fork safety check for Gunicorn multi-worker mode by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2360

• Refact: Add Embedding Token Limit Configuration and Improve Error Handling by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2359

• Refact: Add Embedding Dimension Validation in EmbeddingFunc by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2368

• test: Convert test_workspace_isolation.py to pytest style by @BukeLy in https://github.com/HKUDS/LightRAG/pull/2371

• refactor(chunking): rename params and improve docstring for chunking by @EightyOliveira in https://github.com/HKUDS/LightRAG/pull/2379

• Fix: Add chunk token limit validation with detailed error reporting by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2389

• Fix: Remove redundant exception logging to eliminate pytest shutdown errors by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2390

• issue-2394: use deployment variable instead of model for embeddings API call by @Amrit75 in https://github.com/HKUDS/LightRAG/pull/2395

• Refactor: Centralize keyword_extraction parameter handling in OpenAI LLM implementations by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2401

• Refact: Consolidate Azure OpenAI and OpenAI implementations by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2403

• Update README.md by @chaohuang-ai in https://github.com/HKUDS/LightRAG/pull/2408

• Update README.md by @chaohuang-ai in https://github.com/HKUDS/LightRAG/pull/2409

• feat: create copilot-setup-steps.yml by @netbrah in https://github.com/HKUDS/LightRAG/pull/2410

• Fix: Add Comprehensive Retry Mechanism for Neo4j Storage Operations by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2417

• Refact: Allow API Server to Start Without Built WebUI Assets by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2418

• fix:exception handling order error by @EightyOliveira in https://github.com/HKUDS/LightRAG/pull/2421

• Doc: Update README examples to prevent double-wrapping of embedding functions by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2432

• Fix: Add configurable model support for Jina embedding by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2433

• Fix typos discovered by codespell by @cclauss in https://github.com/HKUDS/LightRAG/pull/2434

• Update README.md by @chaohuang-ai in https://github.com/HKUDS/LightRAG/pull/2439

• Fix KaTeX chemistry formula rendering (\ce command) not working by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2443

• fix(postgres): Add CASCADE to AGE extension creation for automatic dependency resolution by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2446

• Add Python 3.13 and 3.14 to the testing by @cclauss in https://github.com/HKUDS/LightRAG/pull/2436

• Keep GitHub Actions up to date with GitHub's Dependabot by @cclauss in https://github.com/HKUDS/LightRAG/pull/2435

• chore: optimize Dependabot configuration with dependency grouping and PR limits by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2447

• fix: Return existing track_id for duplicate documents in document insertion endpoints by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2469

• feat: Add Cohere reranker config, chunking, and tests by @netbrah in https://github.com/HKUDS/LightRAG/pull/2411

• update to react 19.2.1 by @mccahill in https://github.com/HKUDS/LightRAG/pull/2494

• Fix links in README.md and README-zh.md by @johuellm in https://github.com/HKUDS/LightRAG/pull/2480

• fix(prompt): use language parameter in keywords_extraction prompt by @jcfneto in https://github.com/HKUDS/LightRAG/pull/2491

• feat: Update full-text index name to include workspace label for bett… by @netbrah in https://github.com/HKUDS/LightRAG/pull/2406

• HotFix: MongoDB __post_init__ not called and refactor embedding validation by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2521

• Fix nested EmbeddingFunc wrapping and improve embedding function safety by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2522

• fix(postgres): Handle dollar-sign sequences in Cypher queries to prevent syntax errors by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/2524

New Contributors

• @huangbhan made their first contribution in https://github.com/HKUDS/LightRAG/pull/2327
• @LacombeLouis made their first contribution in https://github.com/HKUDS/LightRAG/pull/2348
• @sleeepyin made their first contribution in https://github.com/HKUDS/LightRAG/pull/2358
• @wmsnp made their first contribution in https://github.com/HKUDS/LightRAG/pull/2378
• @EightyOliveira made their first contribution in https://github.com/HKUDS/LightRAG/pull/2379
• @Amrit75 made their first contribution in https://github.com/HKUDS/LightRAG/pull/2395
• @chaohuang-ai made their first contribution in https://github.com/HKUDS/LightRAG/pull/2408
• @netbrah made their first contribution in https://github.com/HKUDS/LightRAG/pull/2410
• @cclauss made their first contribution in https://github.com/HKUDS/LightRAG/pull/2434
• @mccahill made their first contribution in https://github.com/HKUDS/LightRAG/pull/2494
• @johuellm made their first contribution in https://github.com/HKUDS/LightRAG/pull/2480
• @Ghazi-raad made their first contribution in https://github.com/HKUDS/LightRAG/pull/2426
• @jcfneto made their first contribution in https://github.com/HKUDS/LightRAG/pull/2491

Full Changelog: https://github.com/HKUDS/LightRAG/compare/v1.4.9.8...v1.4.9.9