llama-index

NOTE: Updating to v0.12.0 will require bumping every other llama-index-* package! Every package has had a version bump. Only notable changes are below.

llama-index-core [0.12.0]

• Dropped python3.8 support, Unpinned numpy (#16973)
• Kg/dynamic pg triplet retrieval limit (#16928)

llama-index-indices-managed-llama-cloud [0.6.1]

• Add ID support for LlamaCloudIndex & update from_documents logic, modernize apis (#16927)
• allow skipping waiting for ingestion when uploading file (#16934)
• add support for files endpoints (#16933)

llama-index-indices-managed-vectara [0.3.0]

• Add Custom Prompt Parameter (#16976)

llama-index-llms-bedrock [0.3.0]

• minor fix for messages/completion to prompt (#15729)

llama-index-llms-bedrock-converse [0.4.0]

• Fix async streaming with bedrock converse (#16942)

llama-index-multi-modal-llms-nvidia [0.2.0]

• add vlm support (#16751)

llama-index-readers-confluence [0.3.0]

• Permit passing params to Confluence client (#16961)

llama-index-readers-github [0.5.0]

• Add base URL extraction method to GithubRepositoryReader (#16926)

llama-index-vector-stores-weaviate [1.2.0]

• Allow passing in Weaviate vector store kwargs (#16954)

llama-index-core [0.11.0]

• removed deprecated ServiceContext -- using this now will print an error with a link to the migration guide
• removed deprecated LLMPredictor -- using this now will print an error, any existing LLM is a drop-in replacement
• made pandas an optional dependency
• moved to pydanticV2 officially with full support

Everything Else

• bumped the minor version of every package to account for the new version of llama-index-core

llama-index-core [0.10.68]

• remove nested progress bars in base element node parser (#15550)
• Adding exhaustive docs for workflows (#15556)
• Adding multi-strategy workflow with reflection notebook example (#15445)
• remove openai dep from core (#15527)
• Improve token counter to handle more response types (#15501)
• feat: Allow using step decorator without parentheses (#15540)
• feat: workflow services (aka nested workflows) (#15325)
• Remove requirement to specify "allowed_query_fields" parameter when using "cypher_validator" in TextToCypher retriever (#15506)

llama-index-embeddings-mistralai [0.1.6]

• fix mistral embeddings usage (#15508)

llama-index-embeddings-ollama [0.2.0]

• use ollama client for embeddings (#15478)

llama-index-embeddings-openvino [0.2.1]

• support static input shape for openvino embedding and reranker (#15521)

llama-index-graph-stores-neptune [0.1.8]

• Added code to expose structured schema for Neptune (#15507)

llama-index-llms-ai21 [0.3.2]

• Integration: AI21 Tools support (#15518)

llama-index-llms-bedrock [0.1.13]

• Support token counting for llama-index integration with bedrock (#15491)

llama-index-llms-cohere [0.2.2]

• feat: add tool calling support for achat cohere (#15539)

llama-index-llms-gigachat [0.1.0]

• Adding gigachat LLM support (#15313)

llama-index-llms-openai [0.1.31]

• Fix incorrect type in OpenAI token usage report (#15524)
• allow streaming token counts for openai (#15548)

llama-index-postprocessor-nvidia-rerank [0.2.1]

• add truncate support (#15490)
• Update to 0.2.0, remove old code (#15533)
• update default model to nvidia/nv-rerankqa-mistral-4b-v3 (#15543)

llama-index-readers-bitbucket [0.1.4]

• Fixing the issues in loading file paths from bitbucket (#15311)

llama-index-readers-google [0.3.1]

• enhance google drive reader for improved functionality and usability (#15512)

llama-index-readers-remote [0.1.6]

• check and sanitize remote reader urls (#15494)

llama-index-vector-stores-qdrant [0.2.17]

• fix: setting IDF modifier in QdrantVectorStore for sparse vectors (#15538)

llama-index-core [0.10.59]

• Introduce Workflows for event-driven orchestration (#15067)
• Added feature to context chat engine allowing previous chunks to be inserted into the current context window (#14889)
• MLflow Integration added to docs (#14977)
• docs(literalai): add Literal AI integration to documentation (#15023)
• expand span coverage for query pipeline (#14997)
• make re-raising error skip constructor during asyncio_run() (#14970)

llama-index-embeddings-ollama [0.1.3]

• Add proper async embedding support

llama-index-embeddings-textembed [0.0.1]

• add support for textembed embedding (#14968)

llama-index-graph-stores-falkordb [0.1.5]

• initial implementation FalkorDBPropertyGraphStore (#14936)

llama-index-llms-azure-inference [0.1.1]

• Fix: Azure AI inference integration support for tools (#15044)

llama-index-llms-fireworks [0.1.7]

• Updates to Default model for support for function calling (#15046)

llama-index-llms-ollama [0.2.2]

• toggle for ollama function calling (#14972)
• Add function calling for Ollama (#14948)

llama-index-llms-openllm [0.2.0]

• update to OpenLLM 0.6 (#14935)

llama-index-packs-longrag [0.1.0]

• Adds a LlamaPack that implements LongRAG (#14916)

llama-index-postprocessor-tei-rerank [0.1.0]

• Support for Re-Ranker via Text Embedding Interface (#15063)

llama-index-readers-confluence [0.1.7]

• confluence reader sort auth parameters priority (#14905)

llama-index-readers-file [0.1.31]

• UnstructuredReader use filename as ID (#14946)

llama-index-readers-gitlab [0.1.0]

• Add GitLab reader integration (#15030)

llama-index-readers-google [0.2.11]

• Fix issue with average ratings being a float vs an int (#15070)

llama-index-retrievers-bm25 [0.2.2]

• use proper stemmer in bm25 tokenize (#14965)

llama-index-vector-stores-azureaisearch [0.1.13]

• Fix issue with deleting non-existent index (#14949)

llama-index-vector-stores-elasticsearch [0.2.5]

• disable embeddings for sparse strategy (#15032)

llama-index-vector-stores-kdbai [0.2.0]

• Update default sparse encoder for Hybrid search (#15019)

llama-index-vector-stores-milvus [0.1.22]

• Enhance MilvusVectorStore with flexible index management for overwriting (#15058)

llama-index-vector-stores-postgres [0.1.13]

• Adds option to construct PGVectorStore with a HNSW index (#15024)

llama-index-core [v0.10.57]

• Add an optional parameter similarity_score to VectorContextRetrieve… (#14831)
• add property extraction (using property names and optional descriptions) for KGs (#14707)
• able to attach output classes to LLMs (#14747)
• Add streaming for tool calling / structured extraction (#14759)
• fix from removing private variables when copying/pickling (#14860)
• Fix empty array being send to vector store in ingestion pipeline (#14859)
• optimize ingestion pipeline deduping (#14858)
• Add an optional parameter similarity_score to VectorContextRetriever (#14831)

llama-index-llms-azure-openai [0.1.10]

• Bugfix: AzureOpenAI may fail with custom azure_ad_token_provider (#14869)

llama-index-llms-bedrock-converse [0.1.5]

• feat: ✨ Implement async functionality in BedrockConverse (#14326)

llama-index-llms-langchain [0.3.0]

• make some dependencies optional
• bump langchain version in integration (#14879)

llama-index-llms-ollama [0.1.6]

• Bugfix: ollama streaming response (#14830)

llama-index-multi-modal-llms-anthropic [0.1.5]

• align deps (#14850)

llama-index-readers-notion [0.1.10]

• update notion reader to handle duplicate pages, database+page ids (#14861)

llama-index-vector-stores-milvus [0.1.21]

• Implements delete_nodes() and clear() for Weviate, Opensearch, Milvus, Postgres, and Pinecone Vector Stores (#14800)

llama-index-vector-stores-mongodb [0.1.8]

• MongoDB Atlas Vector Search: Enhanced Metadata Filtering (#14856)

llama-index-vector-stores-opensearch [0.1.13]

• Implements delete_nodes() and clear() for Weviate, Opensearch, Milvus, Postgres, and Pinecone Vector Stores (#14800)

llama-index-vector-stores-pinecone [0.1.8]

• Implements delete_nodes() and clear() for Weviate, Opensearch, Milvus, Postgres, and Pinecone Vector Stores (#14800)

llama-index-vector-stores-postgres [0.1.12]

• Implements delete_nodes() and clear() for Weviate, Opensearch, Milvus, Postgres, and Pinecone Vector Stores (#14800)

llama-index-vector-stores-weaviate [1.0.2]

• Implements delete_nodes() and clear() for Weviate, Opensearch, Milvus, Postgres, and Pinecone Vector Stores (#14800)

llama-index-core [0.10.52]

• fix file reader path bug on windows (#14537)
• follow up with kwargs propagation in colbert index due to change in parent class (#14522)
• deprecate query pipeline agent in favor of FnAgentWorker (#14525O)

llama-index-callbacks-arize-phoenix [0.1.6]

• support latest version of arize #14526

llama-index-embeddings-litellm [0.1.0]

• Add support for LiteLLM Proxy Server for embeddings (#14523)

llama-index-finetuning [0.1.10]

• Adding device choice from sentence_transformers (#14546)

llama-index-graph-stores-neo4 [0.2.7]

• Fixed ordering of returned nodes on vector queries (#14461)

llama-index-llms-bedrock [0.1.10]

• handle empty response in Bedrock AnthropicProvider (#14479)

llama-index-llms-bedrock-converse [0.1.4]

• Fix Bedrock Converse's join_two_dicts function when a new string kwarg is added (#14548)

llama-index-llms-upstage [0.1.4]

• Add upstage tokenizer and token counting method (#14502)

llama-index-readers-azstorage-blob [0.1.7]

• Fix bug with getting object name for blobs (#14547)

llama-index-readers-file [0.1.26]

• Pandas excel reader load data fix for appending documents (#14501)

llama-index-readers-iceberg [0.1.0]

• Add Iceberg Reader integration to LLamaIndex (#14477)

llama-index-readers-notion [0.1.8]

• Added retries (#14488)
• add list_databases method (#14488)

llama-index-readers-slack [0.1.5]

• Enhance SlackReader to fetch Channel IDs from Channel Names/Patterns (#14429)

llama-index-readers-web [0.1.21]

• Add API url to firecrawl reader (#14452)

llama-index-retrievers-bm25 [0.1.5]

• fix score in nodes returned by the BM25 retriever (#14495)

llama-index-vector-stores-azureaisearch [0.1.9]

• add async methods to azure ai search (#14496)

llama-index-vector-stores-kdbai [0.1.8]

• Kdbai rest compatible (#14511)

llama-index-vector-stores-mongodb [0.1.6]

• Adds Hybrid and Full-Text Search to MongoDBAtlasVectorSearch (#14490)