v3.1.0

What's Changed

• minor: Native CosmosTableProvider with namespace partitioning, transactional batch writes, and simplified AzureCosmosStorage.
• patch: Update litellm dependency.

Full Changelog: v3.0.9...3.1.0

v3.0.9

What's Changed

• remove relationships with phantom entities by @dayesouza in #2261
• nlp streaming by @dayesouza in #2264
• Release v3.0.6 by @dayesouza in #2267
• reconfigure vector store size by @dayesouza in #2281
• Pin litellm dependency by @dworthen in #2290
• Release v3.0.7 by @dworthen in #2291
• bump nltk to resolve CVE-2025-14009 by @dworthen in #2298
• Release v3.0.8 by @dworthen in #2299
• Fix broken documentation links. by @dworthen in #2305
• Support client side json validation. by @dworthen in #2306
• parquet reader implementation as a new input type by @gaudyb in #2320
• Release v3.0.9 by @gaudyb in #2321

Full Changelog: v3.0.5...3.0.9

v3.0.8

What's Changed

• patch: bump nltk to resolve CVE-2025-14009

Full Changelog: v3.0.7...v3.0.8

v3.0.7

What's Changed

• patch: reconfigure vector store size by embedding model

Security

• patch: Pin litellm dependency due to litellm compromise: BerriAI/litellm#24512

Full Changelog: v3.0.6...v3.0.7

v3.0.6

What's Changed

• remove relationships with phantom entities by @dayesouza in #2261
• nlp streaming by @dayesouza in #2264

Full Changelog: v3.0.5...v3.0.6

v3.0.5

What's Changed

• fix csv file reader by @dayesouza in #2248
• vector load_documents in batches by @dayesouza in #2251
• Release v3.0.5 by @dayesouza in #2252

Full Changelog: v3.0.4...v3.0.5

v3.0.4

What's Changed

• load_input_documents and create_base_text_units streaming by @dayesouza in #2221
• modify smoke tests to include csv table provider by @dayesouza in #2229
• Cosmosdb communities bug by @gaudyb in #2232
• add release doc by @dayesouza in #2235
• Streaming create communities by @dayesouza in #2237
• Streaming create_final_text_units by @dayesouza in #2238
• generate_text_embeddings streaming by @dayesouza in #2241
• feat(graphrag-vectors): add filtering, timestamps, and CRUD operations by @natoverse in #2236
• streaming finalize_graph by @dayesouza in #2240
• streaming create_final_documents by @dayesouza in #2243
• write stats per workflow by @dayesouza in #2244
• Release v3.0.3 by @gaudyb in #2245
• Release v3.0.4 by @gaudyb in #2247

Full Changelog: v3.0.2...v3.0.4

Release v3.0.2

What's Changed

• patch: Add CSVTableProvider
• patch: Add DataReader class for typed dataframe loading from TableProvider across indexing workflows and query CLI
• patch: Add TableProvider abstraction for table-based storage operations
• patch: Add async iterator support to InputReader and use it in load_input_documents and load_update_documents workflows.
• patch: Add table provider factory.
• patch: Fix missed py 3.13.
• patch: Move document ID, human_readable_id, and raw_data initialization from create_final_documents into load_input_documents and load_update_documents.
• patch: Remove NetworkX dependency from graph utilities; move to DataFrame-based implementations in graphrag.graphs package.
• patch: Remove unnecessary response format check. Fixes: #2203
• patch: add profiling to get memory usage
• patch: update notebooks

Full Changelog: v3.0.1...v3.0.2

Release v3.0.1

What's Changed

• patch: Fix missing dependency.

Full Changelog: https://github.com/microsoft/graphrag/compare/v3.0.0..v3.0.1

Release v3.0.0

What's Changed

• major: Monorepo restructure

  New Packages:

  • graphrag-cache
  • graphrag-chunking
  • graphrag-common
  • graphrag-input
  • graphrag-llm
  • graphrag-storage
  • graphrag-vectors

  Changes:

  • New config: run graphrag init --force to reinitialize config with new layout and options.

Full Changelog: v2.7.1...v3.0.0