Release 0.3.24

What's Changed

• Implement DictPromptTemplate by @nfcampos in #8043
• release(core): 0.3.46 by @benjamincburns in #8052
• fix(tavily): convert search/extract API parameters to snake_case in request body by @abogacki in #8048
• feat(google-common): Gemini 2.5 flash by @afirstenberg in #8051
• feat(core): Add input_variables to serialized representation of DictPromptTemplate by @jacoblee93 in #8055
• release(core): 0.3.47 by @jacoblee93 in #8057
• fix(langchain): Expose _ConfigurableModel in exports by @bracesproul in #8053
• release(langchain): 0.3.24 by @bracesproul in #8059

New Contributors

• @abogacki made their first contribution in #8048

Full Changelog: 0.3.23...0.3.24

Release 0.3.23

What's Changed

• Release 0.3.22 by @jacoblee93 in #8036
• o4 reasoning model support (#8039) by @dudebot in #8040
• release(openai): 0.5.6 by @bracesproul in #8044
• release(langchain): 0.3.23 by @bracesproul in #8045

New Contributors

• @dudebot made their first contribution in #8040

Full Changelog: 0.3.22...0.3.23

langchain==0.3.22

What's Changed

• Release 0.3.21 by @benjamincburns in #7996
• fix(groq): update to latest sdk, format fixes, maintenance by @benjamincburns in #8008
• release(groq): 0.2.2 by @benjamincburns in #8009
• chore(docs): update link to JS JD by @benjamincburns in #8013
• chore(docs): bump nextjs version by @benjamincburns in #8014
• chore: bump release-it version by @benjamincburns in #8015
• chore(community): bump usearch package version by @benjamincburns in #8016
• chore: dedupe yarn.lock, plus required minor code edits by @benjamincburns in #8017
• chore: bump next dep in environment tests by @benjamincburns in #8018
• chore(openai, baidu-qianfan, deepseek, xai): use latest openai SDK & @langchain/openai package by @benjamincburns in #8019
• fix: LangChain tracer should send session name with update requests by @angus-langchain in #8021
• fix: don't drop tool input when converting messages to Anthropic payload by @sd2k in #8025
• release(anthropic): 0.3.18 by @jacoblee93 in #8027
• fix(google-common): Fix warning about system message in Vertex Anthropic by @afirstenberg in #8022
• docs: add GOAT to tool docs by @0xaguspunk in #7906
• feat(community): make convertOpenAPISpecToOpenAIFunctions available for imports by @schipiga in #7955
• fix(langchain): Fix typo in error message for SqlDatabase options by @timler in #7975
• fix(google-genai): Add support for supervisor author name by @sklinkert in #7980
• fix(ci): Remove caching for LangChain CI action by @jacoblee93 in #8028
• docs: Fix urls, remove outdate info in IBM Toolkit by @FilipZmijewski in #7969
• fix(docs): fix parameter casing for useResponsesAPI -> useResponsesApi by @24601 in #8026
• feat(langchain-tavily): creating and implementing the Tavily Partner package for the js library by @michaelgriff in #7957
• release(tavily): 0.1.0 by @benjamincburns in #8029
• chore(tavily): add README.md by @benjamincburns in #8030
• release(tavily): 0.1.1 by @benjamincburns in #8031
• fix(docs): fix broken API ref link, use better link for LangGraph tool calling by @benjamincburns in #8032
• fix(community): perplexity add search domain & recency filter by @anadi45 in #7999
• fix(langchain): Remove extra }} from structured output format instructions by @lx-0 in #8024
• test(google): Gemini 2.5 on Vertex by @afirstenberg in #7927
• docs(community): removed @param tag for non-existent function parameter by @alexTayanovsky in #7997
• release(community): 0.3.41 by @jacoblee93 in #8033
• release(google-genai): 0.2.4 by @jacoblee93 in #8034
• release(core): 0.3.45 by @jacoblee93 in #8035
• Release 0.3.22 by @jacoblee93 in #8036

New Contributors

• @angus-langchain made their first contribution in #8021
• @sd2k made their first contribution in #8025
• @0xaguspunk made their first contribution in #7906
• @schipiga made their first contribution in #7955
• @timler made their first contribution in #7975
• @sklinkert made their first contribution in #7980
• @24601 made their first contribution in #8026
• @michaelgriff made their first contribution in #7957
• @lx-0 made their first contribution in #8024
• @alexTayanovsky made their first contribution in #7997

Full Changelog: 0.3.21...langchain==0.3.22

Release 0.3.21

What's Changed

• Release 0.3.20 by @jacoblee93 in #7958
• fix(openai): ensure that we're not sending stream: true even if streaming is not being done by @dqbd in #7959
• release(openai): 0.5.3 by @dqbd in #7965
• fix(openai): handle system, developer messages by @dqbd in #7966
• release(openai): 0.5.4 by @dqbd in #7967
• fix(docs): fix a typo by @tperelle in #7962
• docs: expand descriptions in llms.txt by @eyurtsev in #7971
• fix(ci): Remove unused debug action by @jacoblee93 in #7976
• docs: fix typo in structured output guide by @zdpk in #7964
• docs: reflect changes in SQL QA prompt by @ccurme in #7985
• chore(docs): update banner to point at JS job listing by @benjamincburns in #7988
• feat: allow defining LangChain-style tools with JSONSchema by @benjamincburns in #7973
• release(core): 0.3.44 by @benjamincburns in #7989
• release(anthropic): 0.3.17 by @benjamincburns in #7990
• release(openai): 0.5.5 by @benjamincburns in #7991
• release(aws): 0.1.8 by @benjamincburns in #7992
• release(cohere): 0.3.3 by @benjamincburns in #7993
• release(google) by @benjamincburns in #7994
• release(community): 0.3.40 by @benjamincburns in #7995

New Contributors

• @tperelle made their first contribution in #7962
• @zdpk made their first contribution in #7964

Full Changelog: 0.3.20...0.3.21

Release 0.3.20

What's Changed

• feat(community,aws): Update @aws-sdk/* dependencies in langchain-aws by @KDKHD in #7710
• fix(core): Fix issue in .d.ts typing for TextEncoder by @danielkatz in #7726
• fix(google-genai): Support larger range of temperatures for Gemini models by @afirstenberg in #7703
• release(aws): 0.1.4 by @jacoblee93 in #7728
• fix(google-common): Eliminate hard-coded default values in favor of model-based defaults by @afirstenberg in #7704
• feat (google-*): Support Google Cloud Express Mode by @afirstenberg in #7676
• minor(google-gauth): Upgrade google-auth-library to latest major version by @afirstenberg in #7729
• fix(google-common): Handle multiple function calls and complex parts/generations by @afirstenberg in #7706
• feat(community): Add support for Amazon Aurora DSQL memory message by @jl4nz in #7635
• release(community): 0.3.32 by @jacoblee93 in #7730
• release(google-vertex): 0.2.0 by @jacoblee93 in #7731
• docs: Fix broken link by @jacoblee93 in #7743
• release(google-genai): Release 0.1.9 by @jacoblee93 in #7748
• feat(anthropic): Support claude 3.7 sonnet & extended thinking by @benjamincburns in #7750
• release(anthropic): 0.3.14 by @benjamincburns in #7751
• fix(pinecone): Update Pinecone SDK to 5.0.0, make peer dep by @jacoblee93 in #7767
• release(pinecone): 0.2.0 by @jacoblee93 in #7772
• docs: Adds missing redirect by @jacoblee93 in #7759
• fix(core): don't create empty text content blocks by @benjamincburns in #7769
• fix(community): usage error returned by Zhipuai streaming data was missing tokens-related information by @mmdapl in #7745
• fix(core): Skip delete operation if not necessary during incremental index by @weakit in #7734
• fix(google-genai): Fix Google Genai usage token by @beowulf11 in #7733
• fix(core): add artifact to ToolMessage and ToolMessageChunk for RemoteRunnable by @acastells in #7763
• fix(deepseek): wrong home link for @langchain/deepseek by @AllenFang in #7754
• release(google-genai): 0.1.10 by @jacoblee93 in #7773
• release(core): 0.3.41 by @jacoblee93 in #7774
• feat(community): Support google cloud storage document loader by @AllenFang in #7740
• release(community): 0.3.33 by @jacoblee93 in #7775
• feat(aws): support reasoning blocks for claude 3.7 by @benjamincburns in #7768
• release(aws): 0.1.5 by @benjamincburns in #7777
• fix(core): Adds fallback for context vars by @jacoblee93 in #7776
• release(core): 0.3.42 by @jacoblee93 in #7783
• fix(anthropic): Log warning if withStructuredOutput is used with thinking models by @jacoblee93 in #7782
• release(anthropic): 0.3.15 by @jacoblee93 in #7784
• fix(aws): fix content block ordering for claude reasoning on bedrock by @benjamincburns in #7785
• release(aws): 0.1.6 by @benjamincburns in #7786
• tests(google-vertex,webauth): Verify code works with gemini-2.0-flash-lite-001 by @afirstenberg in #7794
• Google oauth support for Google calender tool by @deepu105 in #7788
• Google oauth support + fix tools for API v1 by @deepu105 in #7780
• release(community): 0.3.34 by @benjamincburns in #7796
• feat (google-common,vertex,etc): Support Claude 3.7 Thinking via Vertex AI by @afirstenberg in #7793
• release(google-common): 0.2.1 by @benjamincburns in #7799
• release(google-vertexai): 0.2.1 by @benjamincburns in #7800
• release(google-vertexai-web): 0.2.1 by @benjamincburns in #7801
• chore(google-gauth, google-webauth): bump dep version on google-common by @benjamincburns in #7803
• release(google-gauth): 0.2.1 by @benjamincburns in #7804
• release(google-webauth): 0.2.1 by @benjamincburns in #7805
• docs: Fix typos by @rkulyassa in #7823
• fix(aws): fix bedrockRuntimeClinent to refer to endpointHost value by @m-hamaro in #7797
• feature(community): Enable using chat and llms without providing project/space/development id (lightweight engine IBM) by @FilipZmijewski in #7781
• feat(google-common): Add tokenUsage to llmOutput of google-genai by @Liboul in #7806
• docs: broken link in chat models page by @plapinski in #7816
• fix(community): Optional google cloud storage options for GCS doc loader by @AllenFang in #7791
• fix(azure-cosmosdb): Update user agent of AzureCosmosDBMongoDBVectorStore by @fatmelon in #7825
• fix(aws): fix Invalid reasoning content exception on empty reasoning chunk by @ChenNima in #7826
• chore(langchain): add test to ensure sort order on MemoryVectorStore by @benjamincburns in #7792
• release(aws): 0.1.7 by @jacoblee93 in #7836
• release(google): 0.2.2 by @jacoblee93 in #7837
• fix(community): adding structured output parser for Google Calendar by @John2360 in #7789
• release(community): 0.3.35 by @jacoblee93 in #7838
• release(google-genai): 0.1.11 by @jacoblee93 in #7841
• feat(community): add deepseek to allowed bedrock models by @jarib in #7842
• Update sequence.ipynb by @technopahadi in #7847
• feat(community): MariaDB vector store implementation. by @rusher in #7360
• feat(community): Permit bedrock application inference profiles by @bioshazard in #7822
• feat(community): Perplexity integration by @anadi45 in #7817
• release(community): 0.3.36 by @jacoblee93 in #7858
• fix(openai): Fix OpenAI type error by @jacoblee93 in #7857
• release(openai): 0.4.5 by @jacoblee93 in #7859
• fix: Update groq default model by @bracesproul in #7802
• release(groq): 0.2.0 by @bracesproul in #7864
• feat[openai]: responses API by @dqbd in #7831
• release(openai): 0.4.6 by @dqbd in #7869
• fix(openai): Dont populate id field for computer call outputs by @bracesproul in #7871
• release(openai): 0.4.7 by @bracesproul in #7873
• fix(community): Azure AI Search - similaritySearch return description is incorrect by @diberry in #7867
• Revert "fix(azure-cosmosdb): Update user agent of AzureCosmosDBMongoDBVectorStore" by @fatmelon in #7861
• fix(core): Add typing in ChatMessagePromptTemplate.fromTemplate by @gouroujo in #7874
• feat(google-common): Gemma 3 support for AI Studio by @afirstenberg in #7851
• release(azure-cosmosdb): 0.2.8 by @jacoblee93 in #7881
• feat(google-genai): Bump GenAI SDK version by @jacoblee93 in https://github.com/langchain-ai/langch...

Release 0.3.19

What's Changed

• Release 0.3.18 by @jacoblee93 in #7720
• feat(xai): xAI polish by @jacoblee93 in #7722
• feat(langchain): Adds xAI to initChatModel by @jacoblee93 in #7721
• release(xai): 0.0.2 by @bracesproul in #7723
• release(langchain): 0.3.19 by @bracesproul in #7724

Full Changelog: 0.3.18...0.3.19

Release 0.3.18

What's Changed

• Release 0.3.17 by @jacoblee93 in #7715
• docs: Fix the broken link by @toqeer-hussain in #7699
• fix(langchain): Respect split model name for initChatModel by @jacoblee93 in #7716
• fix: typo in the chatbot tutorial by @vedantmishra69 in #7693
• fix(ci): Fix Vercel build by @jacoblee93 in #7717
• fix(community): Fix handling of ChromeAI chunks by @jtpio in #7700
• fix(redis): Add TTL support for redis vector store by @AllenFang in #7695
• fix(redis): update wrong redis setup link by @AllenFang in #7698
• feat(community): Update Voyage embeddings parameters by @nicolas-geysse in #7689
• release(redis): 0.1.1 by @jacoblee93 in #7718
• release(community): 0.3.31 by @jacoblee93 in #7719

New Contributors

• @toqeer-hussain made their first contribution in #7699
• @vedantmishra69 made their first contribution in #7693
• @jtpio made their first contribution in #7700
• @AllenFang made their first contribution in #7695

Full Changelog: 0.3.17...0.3.18

Release 0.3.17

What's Changed

• Release 0.3.16 by @jacoblee93 in #7713
• fix(langchain): Remove stray log by @jacoblee93 in #7714

Full Changelog: 0.3.16...0.3.17

Release 0.3.16

What's Changed

• docs: update README/intro by @vbarda in #7625
• Release 0.3.15 by @jacoblee93 in #7626
• feat(community): improve support for Tavily search tool args by @hugoleborso in #7561
• feat(community): Add boolean metadata type support in Supabase structured query translator by @ArmmGh in #7601
• feat(google-genai): Add support for fileUri in media type in Google GenAI by @hongkongkiwi in #7621
• release(google-genai): 0.1.8 by @jacoblee93 in #7628
• release(community): 0.3.28 by @jacoblee93 in #7629
• fix(docs): imports and call by @crishoj in #7636
• docs: chat model order experiment by @efriis in #7641
• adding interrupt banner by @tanushree-sharma in #7642
• feat(anthropic): Adds Anthropic citations support by @jacoblee93 in #7651
• release(anthropic): 0.3.13 by @jacoblee93 in #7652
• fix(core): Fix coercion of serialized tool and function messages by @jacoblee93 in #7653
• release(core): 0.3.38 by @jacoblee93 in #7654
• weaviate[minor]: Add Document className as metadata by @rolando-cruz in #7486
• release(weaviate): 0.1.1 by @jacoblee93 in #7655
• feat[community]: Add chat deployment to IBM chat class by @FilipZmijewski in #7633
• fix(community): Relax Replicate peer dep by @jacoblee93 in #7656
• feat(community): Add OpenAI Whisper options by @fneiraj in #7650
• release(community): 0.3.29 by @jacoblee93 in #7659
• feat(core): Add a way to manually specify serializable fields by @jacoblee93 in #7667
• release(core): 0.3.39 by @jacoblee93 in #7668
• fix(openai): Prevent extra constructor params from being serialized, add script by @jacoblee93 in #7669
• release(openai): 0.4.3 by @jacoblee93 in #7670
• feat(ollama): Add support for Ollama built-in JSON schema with withStructuredOutput by @jacoblee93 in #7672
• release(ollama): 0.1.6 by @jacoblee93 in #7673
• feat(ollama): Switch Ollama default withStructuredOutput method to jsonSchema by @jacoblee93 in #7681
• release(ollama): 0.2.0 by @jacoblee93 in #7682
• chore(openai): update azure user agent by @sinedied in #7543
• fix(community): Fix interface of chat deployment IBM and add test for this case by @FilipZmijewski in #7666
• docs: Update message_history.ipynb by @yashsharma999 in #7677
• fix(experimental): openai assistant attachments by @zachary-nguyen in #7664
• fix(community): hide console errors on RecursiveUrlLoader by @jfromaniello in #7679
• release(community): 0.3.30 by @jacoblee93 in #7684
• feat(openai): Properly pass through max_completion_tokens by @jacoblee93 in #7683
• release(openai): 0.4.4 by @jacoblee93 in #7685
• feat(core): Expose types/stream as an entrypoint by @jacoblee93 in #7686
• release(core): 0.3.40 by @jacoblee93 in #7687
• feat(langchain): Allow model providers prefixed by colon in initChatModel by @jacoblee93 in #7712

New Contributors

• @hugoleborso made their first contribution in #7561
• @ArmmGh made their first contribution in #7601
• @hongkongkiwi made their first contribution in #7621
• @crishoj made their first contribution in #7636
• @tanushree-sharma made their first contribution in #7642
• @rolando-cruz made their first contribution in #7486
• @fneiraj made their first contribution in #7650
• @yashsharma999 made their first contribution in #7677
• @zachary-nguyen made their first contribution in #7664
• @jfromaniello made their first contribution in #7679

Full Changelog: 0.3.15...0.3.16

Release 0.3.15

What's Changed

• Release 0.3.14 by @jacoblee93 in #7611
• fix: Add test for pdf uploads anthropic by @bracesproul in #7613
• feat: Update google genai to support file uploads by @bracesproul in #7612
• chore(google-genai): Drop .only in test by @bracesproul in #7614
• release(google-genai): 0.1.7 by @bracesproul in #7615
• fix(core): Fix stream events bug when errors are thrown too quickly during iteration by @jacoblee93 in #7617
• release(core): 0.3.37 by @jacoblee93 in #7619
• fix(langchain): Fix Groq import for hub by @jacoblee93 in #7620

Full Changelog: 0.3.14...0.3.15