# LangChain: what changed from 0 to 1

- Product: LangChain (https://whatsnew.fyi/product/langchain)
- Vendor: LangChain
- Range: changelog entries numbered after 0.3.30 up to and including 1.4.0, stable releases only
- Entries below: 25 releases (newest first)
- Resolved: 0 is 0.3.30 and 1 is 1.4.0, the newest stable release of each major we track
- Not shown: 3 releases whose versions could not be placed against this range
- Carrying security changes: 0 · CVEs mentioned: 1 · Mentioning breaking changes: 0 · Removing or deprecating something: 1
- Page: https://whatsnew.fyi/product/langchain/compare/0...1

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

Reuse: the summaries, labels and curation here are © What's New. Quote freely with attribution and a link back; wholesale republication of the corpus is not permitted — terms: https://whatsnew.fyi/terms. The vendors' own release notes remain their publishers'.

## What changed (100 changes, grouped by kind)

### Added

#### 1.4.0 (2026-09-03)

- Add langchain.mcp namespace and MCPAdapter for Model Context Protocol support
- Add runnable langchain.mcp examples to documentation

#### 1.3.16 (2026-08-20)

- Add standard model exception types to core
- Support custom token_counter in ContextEditingMiddleware

#### 1.3.15 (2026-08-11)

- Expose `trace_policy` on `AgentMiddleware`
- Add `state_schema` param to `wrap_tool_call`
- Filter internal middleware model calls from `messages` projection
- Add LangSmith provider to `init_chat_model`
- Add `reasoning_effort` as a standard chat model parameter

#### 1.3.14 (2026-07-16)

- ToolErrorMiddleware

#### 1.3.13 (2026-07-10)

- Add meta extra and support langchain-meta in init_chat_model
- Support explicit prompt caching in OpenAI integration

#### 1.3.7 (2026-06-10)

- Add ProviderToolSearchMiddleware

#### 1.3.5 (2026-06-10)

- Port AND-capable trigger conditions to SummarizationMiddleware
- Support apply_patch built-in tool in openai

#### 1.3.3 (2026-06-02)

- Project subagent runs onto typed run.subagents channel
- Add `interrupt_mode` and `when` predicate to `HumanInTheLoopMiddleware`

#### 1.3.2 (2026-05-26)

- Redact streamed PII in flight on PIIMiddleware
- Register stream transformers on middleware

#### 1.3.0 (2026-05-12)

- Support for version="v3" in stream_events / astream_events for langchain agents

#### 1.3.0a2 (2026-05-06)

- Wire stream_events(version='v3') into create_agent (also in 1.3.0a1)
- Stream_events(version='v3') protocol
- Add `respond` decision to HITL middleware (also in 1.3.0a1; 1.2.17)
- Add content-block-centric streaming (v2) (also in 1.3.0a1)
- `ls_agent_type` tag on `create_agent` calls
- Add LangSmith integration metadata to create_agent and init_chat_model (also in 1.3.0a1)
- Tracing for wrap model + tool call
- Support automatic server-side compaction for OpenAI
- Add langchain-openrouter provider package (also in 1.3.0a1)

#### 1.3.0a1 (2026-05-01)

- Add stream_events(version='v3') protocol
- Add `ls_agent_type` tag on `create_agent` calls
- Add tracing for wrap model and tool call
- Support automatic server-side compaction in OpenAI
- Support state updates from `wrap_model_call` with command(s)

#### 1.2.16 (2026-04-29)

- Add content-block-centric streaming (v2) in core
- Add ls_agent_type tag on create_agent calls

### Changed

#### 1.4.0 (2026-09-03)

- Omit middleware trace inputs in anthropic and langchain for performance improvement
- Bump vcrpy test dependency minimum to version 8.2.0

#### 1.3.11 (2026-06-22)

- Bump pydantic-settings from 2.12.0 to 2.14.2
- Bump vcrpy from 8.1.1 to 8.2.1
- Bump langsmith from 0.8.9 to 0.8.18

#### 1.3.10 (2026-06-18)

- Bump cryptography from 46.0.7 to 48.0.1
- Bump aiohttp from 3.14.0 to 3.14.1
- Bump pyjwt from 2.12.0 to 2.13.0

#### 1.3.8 (2026-06-12)

- Add overloads to create_agent

#### 1.3.7 (2026-06-10)

- Activate mypy warn_return_any rule

#### 1.3.3 (2026-06-02)

- Bump `langgraph` to 1.2.4
- Loosen `langgraph` dependency range

#### 1.3.2 (2026-05-26)

- Require langgraph>=1.2.2

#### 1.2.18 (2026-05-08)

- Retarget deprecations to create_agent

#### 1.3.0a2 (2026-05-06)

- Stop inlining agent state into tool-dispatch `Send`s (also in 1.3.0a1)
- Add lru cache when resolving types
- Reduce init speed by 15%

#### 1.2.16 (2026-04-29)

- Stop inlining agent state into tool-dispatch Sends to improve performance
- Add LRU cache when resolving types for improved performance

### Fixed

#### 1.4.0 (2026-09-03)

- Include model destination in agent tool routing

#### 1.3.18 (2026-08-27)

- Preserve content-block shape in PIIMiddleware redaction
- Shore up indexing in genai v1 streaming content

#### 1.3.17 (2026-08-25)

- Frame custom HITL rejection reasons

#### 1.3.16 (2026-08-20)

- Re-raise non-retryable exceptions in ModelRetryMiddleware
- Preserve final repeated schema ordering

#### 1.3.15 (2026-08-11)

- Preserve history on `SummarizationMiddleware` summary failure
- Handle import error in `LLMToolEmulator` by `model`
- Re-export `PIIMatch` from `middleware` package
- Restrict narrowed `ToolStrategy` in bound tools
- Handle malformed structured-output responses
- Prevent orphaned `tool_calls` in `ToolCallLimitMiddleware` end behavior
- Add aliases for bedrock mantle chat models
- Clear stale `structured_response` between checkpointed turns
- Stop HITL approval gates from silently failing open
- Propagate model middleware control flow

#### 1.3.14 (2026-07-16)

- Only retry retryable exceptions in ToolRetryMiddleware

#### 1.3.12 (2026-07-08)

- Propagate interrupts through ToolRetryMiddleware
- Avoid shared process-group kill in shell middleware
- Sanitize anthropic cache markers on fallback retries

#### 1.3.11 (2026-06-22)

- Only set `strict=True` on tools for OpenAI-compatible models in `ProviderStrategy`

#### 1.3.10 (2026-06-18)

- Switch summary format
- Detect provider strategy for dated gpt-5.2/gpt-5.4 snapshots

#### 1.3.9 (2026-06-12)

- Confine file-search results and tighten anthropic allowed_prefixes

#### 1.3.8 (2026-06-12)

- Support async middleware decorator typing
- Tighten structured output model fallbacks

#### 1.3.6 (2026-06-10)

- Preserve summarization trigger compatibility

#### 1.3.5 (2026-06-10)

- Fix minimum core dependency requirement for openai

#### 1.3.4 (2026-06-02)

- Improve HITL rejection guidance

#### 1.3.2 (2026-05-26)

- Land final answer in last AIMessage for TodoListMiddleware
- Sort glob_search results by mtime (newest first)

#### 1.3.1 (2026-05-15)

- Alias Bedrock providers in summarization token check

#### 1.3.0a2 (2026-05-06)

- Ordered schema resolution — list replaces set so state_schema wins
- Update recursion limit for create_agent (also in 1.3.0a1)
- Recognize ChatAnthropicVertex in _get_approximate_token_counter (also in 1.3.0a1)
- Export Runtime from agents.middleware (also in 1.3.0a1)
- Normalize custom detector output to prevent KeyError in hash/mask strategies (also in 1.3.0a1)
- Support anthropic-bedrock in init_chat_model (also in 1.3.0a1)
- Fix token counting on partial message sequences (also in 1.3.0a1)
- Normalize raw schemas in middleware response_format override (also in 1.3.0a1)

#### 1.3.0a1 (2026-05-01)

- Honor `max_retries` in Fireworks
- Allow Gemini 3 models to use `ProviderStrategy` with tools

#### 1.2.16 (2026-04-29)

- Honor max_retries parameter in Fireworks integration

### Removed

#### 1.2.18 (2026-05-08)

- Remove ls_agent_type tag from create_agent calls

### Deprecated

#### 1.2.18 (2026-05-08)

- Deprecate hub and limit loads/dumps in langchain-classic

## Release notes

### 1.4.0

- Date: 2026-09-03
- Version: 1.4.0
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.4.0
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.4.0

- **added** — Add langchain.mcp namespace and MCPAdapter for Model Context Protocol support
- **added** — Add runnable langchain.mcp examples to documentation
- **changed** — Omit middleware trace inputs in anthropic and langchain for performance improvement
- **fixed** — Include model destination in agent tool routing
- **changed** — Bump vcrpy test dependency minimum to version 8.2.0

Changes since langchain==1.3.18

docs(langchain): runnable `langchain.mcp` examples (#39976)
feat(langchain): `langchain.mcp` namespace, `MCPAdapter` (#39939)
perf(anthropic,langchain): omit middleware trace inputs (#40098)
fix(langchain): include model destination in agent tool routing (#38355)
chore(langchain): bump vcrpy test dependency minimum to `>=8.2.0` (#39942)

### 1.3.18

- Date: 2026-08-27
- Version: 1.3.18
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.18
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.18

- **fixed** — Preserve content-block shape in PIIMiddleware redaction
- **fixed** — Shore up indexing in genai v1 streaming content

Changes since langchain==1.3.17

release(langchain): 1.3.18 (#39966)
fix(langchain): preserve content-block shape in PIIMiddleware redaction (#39894)
fix(core): shore up indexing in genai v1 streaming content (#39964)

### 1.3.17

- Date: 2026-08-25
- Version: 1.3.17
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.17
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.17

- **fixed** — Frame custom HITL rejection reasons

Changes since langchain==1.3.16

release(langchain): 1.3.17 (#39893)
fix(langchain): frame custom HITL rejection reasons (#39773)
chore(deps): bump minor and patch dependencies (#39869)

### 1.3.16

- Date: 2026-08-20
- Version: 1.3.16
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.16
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.16

- **added** — Add standard model exception types to core
- **added** — Support custom token_counter in ContextEditingMiddleware
- **fixed** — Re-raise non-retryable exceptions in ModelRetryMiddleware
- **fixed** — Preserve final repeated schema ordering

Changes since langchain==1.3.15

release(langchain): 1.3.16 (#39806)
feat(core): add standard model exception types (#39538)
feat(langchain): support custom token_counter in ContextEditingMiddleware (#39754)
fix(langchain): re-raise non-retryable exceptions in ModelRetryMiddleware (#38960)
chore(langchain): update docs on error handling for json schema (#39632)
fix(langchain): preserve final repeated schema ordering (#39284)

### 1.3.15

- Date: 2026-08-11
- Version: 1.3.15
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.15
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.15

- **added** — Expose `trace_policy` on `AgentMiddleware`
- **fixed** — Preserve history on `SummarizationMiddleware` summary failure
- **fixed** — Handle import error in `LLMToolEmulator` by `model`
- **added** — Add `state_schema` param to `wrap_tool_call`
- **fixed** — Re-export `PIIMatch` from `middleware` package
- **fixed** — Restrict narrowed `ToolStrategy` in bound tools
- **added** — Filter internal middleware model calls from `messages` projection
- **fixed** — Handle malformed structured-output responses
- **fixed** — Prevent orphaned `tool_calls` in `ToolCallLimitMiddleware` end behavior
- **fixed** — Add aliases for bedrock mantle chat models
- **added** — Add LangSmith provider to `init_chat_model`
- **fixed** — Clear stale `structured_response` between checkpointed turns
- **fixed** — Stop HITL approval gates from silently failing open
- **fixed** — Propagate model middleware control flow
- **added** — Add `reasoning_effort` as a standard chat model parameter

Changes since langchain==1.3.14

release(langchain): 1.3.15 (#39595)
feat(langchain): expose `trace_policy` on `AgentMiddleware` (#38910)
chore(langchain): fix type errors in tests (#39589)
chore: bump h2 from 4.3.0 to 4.4.1 in /libs/langchain_v1 (#39324)
fix(langchain): preserve history on `SummarizationMiddleware` summary failure (#39268)
fix(langchain): handle import error in `LLMToolEmulator` by `model` (#39290)
refactor(langchain): update doc strings (#39305)
feat(langchain): add `state_schema` param to `wrap_tool_call` (#39292)
fix(langchain): re-export `PIIMatch` from `middleware` package (#39291)
fix(langchain): restrict narrowed `ToolStrategy` in bound tools (#39259)
feat(langchain): filter internal middleware model calls from `messages` projection (#39252)
chore: bump aiohttp from 3.14.1 to 3.14.3 in /libs/langchain_v1 (#39242)
chore: bump cryptography from 48.0.1 to 50.0.0 in /libs/langchain_v1 (#39240)
test(langchain): regression test for shell tool + checkpointer msgpack error (#39267)
fix(langchain): handle malformed structured-output responses (#39245)
fix(langchain): prevent orphaned `tool_calls` in `ToolCallLimitMiddleware` end behavior (#39258)
fix(langchain): add aliases for bedrock mantle chat models (#39260)
feat(langchain): add LangSmith provider to `init_chat_model` (#39224)
fix(langchain): clear stale `structured_response` between checkpointed turns (#39248)
fix(langchain): stop HITL approval gates from silently failing open (#39247)
chore: bump the minor-and-patch group across 3 directories with 7 updates (#39187)
fix(langchain): propagate model middleware control flow (#39199)
fix(langchain): update Anthropic config test (#39172)
chore: bump pyasn1 from 0.6.3 to 0.6.4 in /libs/langchain_v1 (#39026)
chore: bump pillow from 12.2.0 to 12.3.0 in /libs/langchain_v1 (#38994)
feat(core): add `reasoning_effort` as a standard chat model parameter (#38887)
docs(langchain): clarify `ToolRetryMiddleware` exception handling (#38884)

### 1.3.14

- Date: 2026-07-16
- Version: 1.3.14
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.14
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.14

- **fixed** — Only retry retryable exceptions in ToolRetryMiddleware
- **added** — ToolErrorMiddleware

Changes since langchain==1.3.13

release(langchain): 1.3.14 (#38883)
fix(langchain): only retry retryable exceptions in `ToolRetryMiddleware` (#38845)
feat(langchain): `ToolErrorMiddleware` (#38781)

### 1.3.13

- Date: 2026-07-10
- Version: 1.3.13
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.13
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.13

- **added** — Add meta extra and support langchain-meta in init_chat_model
- **added** — Support explicit prompt caching in OpenAI integration

Changes since langchain==1.3.12

release(langchain): 1.3.13 (#38787)
feat(langchain): add `meta` extra and support langchain-meta in init_chat_model (#38786)
feat(openai): support explicit prompt caching (#38762)
chore(deps): refresh lockfiles (#38746)

### 1.3.12

- Date: 2026-07-08
- Version: 1.3.12
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.12
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.12

- **fixed** — Propagate interrupts through ToolRetryMiddleware
- **fixed** — Avoid shared process-group kill in shell middleware
- **fixed** — Sanitize anthropic cache markers on fallback retries

Changes since langchain==1.3.11

release(langchain): 1.3.12 (#38730)
fix(langchain): propagate interrupts through ToolRetryMiddleware (#38722)
fix(langchain): avoid shared process-group kill in shell middleware (#36359)
fix(langchain): sanitize anthropic cache markers on fallback retries (#37867)
style: fix some ruff preview rules in `langchain_v1` and `standard-tests` (#38657)
chore(langchain): add types in agent middleware tests (#38188)

### 1.3.11

- Date: 2026-06-22
- Version: 1.3.11
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.11
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.11

- **fixed** — Only set `strict=True` on tools for OpenAI-compatible models in `ProviderStrategy`
- **changed** — Bump pydantic-settings from 2.12.0 to 2.14.2
- **changed** — Bump vcrpy from 8.1.1 to 8.2.1
- **changed** — Bump langsmith from 0.8.9 to 0.8.18

Changes since langchain==1.3.10

release(langchain): 1.3.11 (#38377)
fix(langchain,openai): only set `strict=True` on tools for OpenAI-compatible models in `ProviderStrategy` (#38370)
chore: bump pydantic-settings from 2.12.0 to 2.14.2 in /libs/langchain_v1 (#38279)
chore: bump vcrpy from 8.1.1 to 8.2.1 in /libs/langchain_v1 (#38280)
chore: bump langsmith from 0.8.9 to 0.8.18 in /libs/langchain_v1 (#38281)
docs(langchain): document summarization prompt contract (#38256)

### 1.3.10

- Date: 2026-06-18
- Version: 1.3.10
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.10
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.10

- **fixed** — Switch summary format
- **fixed** — Detect provider strategy for dated gpt-5.2/gpt-5.4 snapshots
- **changed** — Bump cryptography from 46.0.7 to 48.0.1
- **changed** — Bump aiohttp from 3.14.0 to 3.14.1
- **changed** — Bump pyjwt from 2.12.0 to 2.13.0

Changes since langchain==1.3.9

release(langchain): 1.3.10 (#38255)
chore: bump cryptography from 46.0.7 to 48.0.1 in /libs/langchain_v1 (#38176)
chore: bump aiohttp from 3.14.0 to 3.14.1 in /libs/langchain_v1 (#38179)
fix(langchain): switch summary format (#38171)
fix(langchain): detect provider strategy for dated `gpt-5.2`/`gpt-5.4` snapshots (#38222)
chore(langchain): improve typing in tests (#38163)
chore: bump pyjwt from 2.12.0 to 2.13.0 in /libs/langchain_v1 (#38168)
release(openai): 1.3.2 (#38130)
hotfix(openai): switch version (#38123)
release(openai): 1.4.0 (#38120)
docs: refresh `README` installation and resources (#38119)
test(core,langchain): update tests for explicit deserialization allowlists (#38118)
release(core): 1.4.7 (#38111)
release(anthropic): 1.4.6 (#38105)

### 1.3.9

- Date: 2026-06-12
- Version: 1.3.9
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.9
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.9

- **fixed** — Confine file-search results and tighten anthropic allowed_prefixes

Changes since langchain==1.3.8

release(anthropic): 1.4.6 (#38105)
release(langchain): 1.3.9 (#38104)
fix(langchain,anthropic): confine file-search results and tighten anthropic `allowed_prefixes` (#38106)

### 1.3.8

- Date: 2026-06-12
- Version: 1.3.8
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.8
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.8

- **fixed** — Support async middleware decorator typing
- **fixed** — Tighten structured output model fallbacks
- **changed** — Add overloads to create_agent

Changes since langchain==1.3.7

release(langchain): 1.3.8 (#38096)
style(core,langchain,langchain-classic,partners): replace double backticks in docstrings (#38095)
release(core): 1.4.6 (#38061)
chore(langchain): add overloads to `create_agent` (#34309)
chore(infra): bump mypy to 2.1 and unify type-check config across the monorepo (#36470)
fix(langchain): support async middleware decorator typing (#34584)
fix(langchain): tighten structured output model fallbacks (#38042)
release(anthropic): 1.4.5 (#38036)
hotfix(core): bump lockfile(s) (#38032)
refactor(langchain): refactor `test_create_agent_tool_validation` (#34443)

### 1.3.7

- Date: 2026-06-10
- Version: 1.3.7
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.7
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.7

- **added** — Add ProviderToolSearchMiddleware
- **changed** — Activate mypy warn_return_any rule

Changes since langchain==1.3.6

release(langchain): 1.3.7 (#38024)
style(langchain): add ruff rules ARG (#34435)
feat(langchain): add `ProviderToolSearchMiddleware` (#37969)
chore(langchain): activate mypy `warn_return_any` (#34249)
test(langchain): mark legacy trigger view for 2.0 removal (#38002)

### 1.3.6

- Date: 2026-06-10
- Version: 1.3.6
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.6
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.6

- **fixed** — Preserve summarization trigger compatibility

Changes since langchain==1.3.5

release(langchain): 1.3.6 (#38001)
fix(langchain): preserve summarization trigger compatibility (#38000)

### 1.3.5

- Date: 2026-06-10
- Version: 1.3.5
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.5
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.5

- **added** — Port AND-capable trigger conditions to SummarizationMiddleware
- **added** — Support apply_patch built-in tool in openai
- **fixed** — Fix minimum core dependency requirement for openai

Changes since langchain==1.3.4

release(langchain): 1.3.5 (#37998)
feat(langchain): port AND-capable trigger conditions to `SummarizationMiddleware` (#34576)
hotfix(openai): min core dep (#37990)
feat(openai): support `apply_patch` built-in tool (#37157)
chore: bump pyarrow from 21.0.0 to 23.0.1 in /libs/langchain_v1 (#37930)
chore: bump dependencies  (#37892)
chore: bump aiohttp from 3.13.4 to 3.14.0 in /libs/langchain_v1 (#37888)

### 1.3.4

- Date: 2026-06-02
- Version: 1.3.4
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.4
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.4

- **fixed** — Improve HITL rejection guidance

Changes since langchain==1.3.3

release(langchain): 1.3.4 (#37861)
fix(langchain): improve HITL rejection guidance (#37859)

### 1.3.3

- Date: 2026-06-02
- Version: 1.3.3
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.3
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.3

- **added** — Project subagent runs onto typed run.subagents channel
- **added** — Add `interrupt_mode` and `when` predicate to `HumanInTheLoopMiddleware`
- **changed** — Bump `langgraph` to 1.2.4
- **changed** — Loosen `langgraph` dependency range

Changes since langchain==1.3.2

release(langchain): 1.3.3 (#37843)
chore(langchain): bump `langgraph` to 1.2.4 (#37857)
chore(langchain): loosen `langgraph` dep range (#37855)
feat(langchain): project subagent runs onto typed run.subagents channel (#37739)
feat(langchain): add `interrupt_mode` and `when` predicate to `HumanInTheLoopMiddleware` (#37579)

### 1.3.2

- Date: 2026-05-26
- Version: 1.3.2
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.2
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.2

- **changed** — Require langgraph>=1.2.2
- **fixed** — Land final answer in last AIMessage for TodoListMiddleware
- **added** — Redact streamed PII in flight on PIIMiddleware
- **added** — Register stream transformers on middleware
- **fixed** — Sort glob_search results by mtime (newest first)

Changes since langchain==1.3.1

chore(langchain): bump to 1.3.2, require langgraph>=1.2.2 (#37703)
fix(langchain): land final answer in last AIMessage for `TodoListMiddleware` (#37643)
feat(langchain): redact streamed PII in flight on `PIIMiddleware` (#37616)
chore(infra): bump `langchain-tests` floor to 1.1.9 (#37610)
release(standard-tests): 1.1.9 (#37609)
feat(langchain): register stream transformers on middleware (#37591)
chore: bump idna from 3.11 to 3.15 in /libs/langchain_v1 (#37534)
ci(infra): harden Dependabot version-bound preservation (#37510)
refactor(langchain): clean redundancy ifelse. (#37448)
hotfix: bump lockfiles (#37508)
chore(core,langchain,openai): refresh stale OpenAI model references (#37487)
fix(langchain): sort glob_search results by mtime (newest first) (#37462)
chore: bump langsmith from 0.7.31 to 0.8.0 in /libs/langchain_v1 (#37391)

### 1.3.1

- Date: 2026-05-15
- Version: 1.3.1
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.1
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.1

- **fixed** — Alias Bedrock providers in summarization token check

Changes since langchain==1.3.0

release(langchain): 1.3.1 (#37454)
fix(langchain): alias Bedrock providers in summarization token check (#37453)

### 1.3.0

- Date: 2026-05-12
- Version: 1.3.0
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.0
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.0

- **added** — Support for version="v3" in stream_events / astream_events for langchain agents

This release adds support for `version="v3"` in `stream_events` / `astream_events` for `langchain` agents. Refer to the [event streaming](https://docs.langchain.com/oss/python/langchain/event-streaming) guide for details.

### 1.2.18

- Date: 2026-05-08
- Version: 1.2.18
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.2.18
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.2.18

- **removed** — Remove ls_agent_type tag from create_agent calls
- **deprecated** — Deprecate hub and limit loads/dumps in langchain-classic
- **changed** — Retarget deprecations to create_agent

Changes since langchain==1.2.17

release(langchain): 1.2.18 (#37250)
revert: feat(langchain): ls_agent_type tag on create_agent calls (#37249)
chore(langchain-classic): deprecate hub, limit loads/dumps (#37234)
refactor(langchain-classic): retarget deprecations to `create_agent`, other chores (#37164)
chore(langchain,langchain-classic): uncomment optional deps (#37163)
release(fireworks): 1.2.1 (#37113)

### 1.3.0a2

- Date: 2026-05-06
- Version: 1.3.0a2
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.0a2
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.0a2

- **fixed** — Ordered schema resolution — list replaces set so state_schema wins
- **added** — Wire stream_events(version='v3') into create_agent
- **added** — Stream_events(version='v3') protocol
- **added** — Add `respond` decision to HITL middleware
- **changed** — Stop inlining agent state into tool-dispatch `Send`s
- **added** — Add content-block-centric streaming (v2)
- **added** — `ls_agent_type` tag on `create_agent` calls
- **changed** — Add lru cache when resolving types
- **changed** — Reduce init speed by 15%
- **fixed** — Update recursion limit for create_agent
- **fixed** — Recognize ChatAnthropicVertex in _get_approximate_token_counter
- **added** — Add LangSmith integration metadata to create_agent and init_chat_model
- **fixed** — Export Runtime from agents.middleware
- **added** — Tracing for wrap model + tool call
- **fixed** — Normalize custom detector output to prevent KeyError in hash/mask strategies
- **fixed** — Support anthropic-bedrock in init_chat_model
- **added** — Support automatic server-side compaction for OpenAI
- **added** — Add langchain-openrouter provider package
- **fixed** — Fix token counting on partial message sequences
- **fixed** — Normalize raw schemas in middleware response_format override

Initial release

release(langchain): 1.3.0a2 (#37225)
release(langchain): 1.3.0a2 (#37224)
fix(langchain): ordered schema resolution — list replaces set so state_schema wins (#37223)
release(langchain): 1.3.0a1 (#37140)
feat(langchain): wire stream_events(version='v3') into create_agent (#37136)
Merge remote-tracking branch 'origin/master' into v1.4
feat(core): stream_events(version='v3') protocol (#37111)
release(fireworks): 1.2.1 (#37113)
release(langchain): 1.2.17 (#37110)
feat(langchain): add `respond` decision to HITL middleware (#37095)
release(langchain): 1.2.16 (#37085)
chore(docs): update x handle references (#37081)
release(anthropic): 1.4.2 (#37061)
perf(langchain): stop inlining agent state into tool-dispatch `Send`s (#36960)
docs(langchain): correct import paths in agent middleware docstrings (#36987)
hotfix: bump min core versions (#36996)
release(openai): 1.2.1 (#36995)
feat(core): add content-block-centric streaming (v2) (#36834)
fix(fireworks): honor `max_retries` (#36973)
release(openai): 1.2.0 (#36961)
docs(langchain): correct import path in `ModelCallLimitMiddleware` docstring (#36895)
feat(langchain): `ls_agent_type` tag on `create_agent` calls (#36774)
release(core): release 1.3.0 (#36851)
chore: bump langsmith from 0.6.3 to 0.7.31 in /libs/langchain_v1 (#36800)
chore(deps): bump pytest to `9.0.3` (#36801)
chore: bump pytest from 9.0.2 to 9.0.3 in /libs/langchain_v1 (#36717)
chore: bump pillow from 12.1.1 to 12.2.0 in /libs/langchain_v1 (#36710)
test(langchain): test cache hit/miss on lru (#36659)
perf(langchain): add lru cache when resolving types (#36649)
perf(langchain): add another init test with middleware (#36644)
chore(core): reduce streaming metadata / perf (#36588)
perf(langchain): add benchmark command (#36641)
chore: bump cryptography from 46.0.6 to 46.0.7 in /libs/langchain_v1 (#36619)
docs(langchain): add missing baseten and litellm to `init_chat_model` (#36562)
release(core): 1.2.26 (#36511)
release: langchain v1.2.15 (#36496)
chore: bump aiohttp from 3.13.3 to 3.13.4 in /libs/langchain_v1 (#36438)
release(langchain): 1.2.14 (#36396)
chore: pygments>=2.20.0 across all packages (CVE-2026-4539) (#36385)
test(langchain): cover runtime recursion limit override in create_agent (#36376)
perf(langchain): reduce init speed by 15% (#36375)
fix(langchain): update recursion limit for create_agent (#36351)
fix(infra): correct lint_diff relative paths in package makefiles (#36333)
chore: bump cryptography from 46.0.5 to 46.0.6 in /libs/langchain_v1 (#36324)
fix(langchain): recognize ChatAnthropicVertex in _get_approximate_token_counter (#36320)
chore(langchain): remove unnecessary description for toods list as a group (#36315)
chore(langchain): add async implementation to todolist and test (#36313)
chore(langchain): speed up todo list middleware init (#36311)
chore: bump requests from 2.32.5 to 2.33.0 in /libs/langchain_v1 (#36241)
chore(partners): bump `langchain-core` min to `1.2.21` (#36183)
fix(core,model-profiles): add missing `ModelProfile` fields, warn on schema drift (#36129)
fix(langchain,langchain-classic): update model provider classes for Azure AI Foundry (#35812)
release(langchain): 1.2.13 (#36111)
ci: suppress pytest streaming output in CI (#36092)
feat: Add LangSmith integration metadata to create_agent and init_chat_model (#35810)
ci: avoid unnecessary dep installs in lint targets (#36046)
chore: bump pyasn1 from 0.6.2 to 0.6.3 in /libs/langchain_v1 (#36029)
chore: bump pyjwt from 2.10.1 to 2.12.0 in /libs/langchain_v1 (#36024)
fix(langchain): export Runtime from agents.middleware (#35975)
chore: bump locks, lint (#35985)
fix(openai): add type: message to Responses API input items (#35693)
chore: bump black from 25.11.0 to 26.3.1 in /libs/langchain_v1 (#35802)
chore: bump orjson from 3.11.5 to 3.11.6 in /libs/langchain_v1 (#35807)
chore(langchain): add baseten to `_BUILTIN_PROVIDERS` (#35777)
release(langchain_v1): 1.2.12 (#35770)
feat: tracing for wrap model + tool call (#35765)
release(

_[Truncated at 4000 characters — full notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.0a2]_

### 1.3.0a1

- Date: 2026-05-01
- Version: 1.3.0a1
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.0a1
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.3.0a1

- **added** — Wire stream_events(version='v3') into create_agent
- **added** — Add stream_events(version='v3') protocol
- **added** — Add `respond` decision to HITL middleware
- **added** — Add content-block-centric streaming (v2)
- **added** — Add `ls_agent_type` tag on `create_agent` calls
- **added** — Add LangSmith integration metadata to create_agent and init_chat_model
- **added** — Add tracing for wrap model and tool call
- **added** — Add langchain-openrouter provider package
- **added** — Support automatic server-side compaction in OpenAI
- **added** — Support state updates from `wrap_model_call` with command(s)
- **fixed** — Honor `max_retries` in Fireworks
- **fixed** — Update recursion limit for create_agent
- **fixed** — Recognize ChatAnthropicVertex in _get_approximate_token_counter
- **fixed** — Export Runtime from agents.middleware
- **fixed** — Normalize custom detector output to prevent KeyError in hash/mask strategies
- **fixed** — Support anthropic-bedrock in init_chat_model
- **fixed** — Allow Gemini 3 models to use `ProviderStrategy` with tools
- **fixed** — Fix token counting on partial message sequences
- **fixed** — Normalize raw schemas in middleware response_format override
- **changed** — Stop inlining agent state into tool-dispatch `Send`s

Initial release

release(langchain): 1.3.0a1 (#37140)
feat(langchain): wire stream_events(version='v3') into create_agent (#37136)
Merge remote-tracking branch 'origin/master' into v1.4
feat(core): stream_events(version='v3') protocol (#37111)
release(fireworks): 1.2.1 (#37113)
release(langchain): 1.2.17 (#37110)
feat(langchain): add `respond` decision to HITL middleware (#37095)
release(langchain): 1.2.16 (#37085)
chore(docs): update x handle references (#37081)
release(anthropic): 1.4.2 (#37061)
perf(langchain): stop inlining agent state into tool-dispatch `Send`s (#36960)
docs(langchain): correct import paths in agent middleware docstrings (#36987)
hotfix: bump min core versions (#36996)
release(openai): 1.2.1 (#36995)
feat(core): add content-block-centric streaming (v2) (#36834)
fix(fireworks): honor `max_retries` (#36973)
release(openai): 1.2.0 (#36961)
docs(langchain): correct import path in `ModelCallLimitMiddleware` docstring (#36895)
feat(langchain): `ls_agent_type` tag on `create_agent` calls (#36774)
release(core): release 1.3.0 (#36851)
chore: bump langsmith from 0.6.3 to 0.7.31 in /libs/langchain_v1 (#36800)
chore(deps): bump pytest to `9.0.3` (#36801)
chore: bump pytest from 9.0.2 to 9.0.3 in /libs/langchain_v1 (#36717)
chore: bump pillow from 12.1.1 to 12.2.0 in /libs/langchain_v1 (#36710)
test(langchain): test cache hit/miss on lru (#36659)
perf(langchain): add lru cache when resolving types (#36649)
perf(langchain): add another init test with middleware (#36644)
chore(core): reduce streaming metadata / perf (#36588)
perf(langchain): add benchmark command (#36641)
chore: bump cryptography from 46.0.6 to 46.0.7 in /libs/langchain_v1 (#36619)
docs(langchain): add missing baseten and litellm to `init_chat_model` (#36562)
release(core): 1.2.26 (#36511)
release: langchain v1.2.15 (#36496)
chore: bump aiohttp from 3.13.3 to 3.13.4 in /libs/langchain_v1 (#36438)
release(langchain): 1.2.14 (#36396)
chore: pygments>=2.20.0 across all packages (CVE-2026-4539) (#36385)
test(langchain): cover runtime recursion limit override in create_agent (#36376)
perf(langchain): reduce init speed by 15% (#36375)
fix(langchain): update recursion limit for create_agent (#36351)
fix(infra): correct lint_diff relative paths in package makefiles (#36333)
chore: bump cryptography from 46.0.5 to 46.0.6 in /libs/langchain_v1 (#36324)
fix(langchain): recognize ChatAnthropicVertex in _get_approximate_token_counter (#36320)
chore(langchain): remove unnecessary description for toods list as a group (#36315)
chore(langchain): add async implementation to todolist and test (#36313)
chore(langchain): speed up todo list middleware init (#36311)
chore: bump requests from 2.32.5 to 2.33.0 in /libs/langchain_v1 (#36241)
chore(partners): bump `langchain-core` min to `1.2.21` (#36183)
fix(core,model-profiles): add missing `ModelProfile` fields, warn on schema drift (#36129)
fix(langchain,langchain-classic): update model provider classes for Azure AI Foundry (#35812)
release(langchain): 1.2.13 (#36111)
ci: suppress pytest streaming output in CI (#36092)
feat: Add LangSmith integration metadata to create_agent and init_chat_model (#35810)
ci: avoid unnecessary dep installs in lint targets (#36046)
chore: bump pyasn1 from 0.6.2 to 0.6.3 in /libs/langchain_v1 (#36029)
chore: bump pyjwt from 2.10.1 to 2.12.0 in /libs/langchain_v1 (#36024)
fix(langchain): export Runtime from agents.middleware (#35975)
chore: bump locks, lint (#35985)
fix(openai): add type: message to Responses API input items (#35693)
chore: bump black from 25.11.0 to 26.3.1 in /libs/langchain_v1 (#35802)
chore: bump orjson from 3.11.5 to 3.11.6 in /libs/langchain_v1 (#35807)
chore(langchain): add baseten to `_BUILTIN_PROVIDERS` (#35777)
release(langchain_v1): 1.2.12 (#35770)
feat: tracing for wrap model + tool call (#35765)
release(langchain_v1): 1.2.11 (#35723)
fix(langchain): normalize custom detector output to prevent KeyError in hash/mask strategies (#35651)
chore: bump the minor-and-patch g

_[Truncated at 4000 characters — full notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.3.0a1]_

### 1.2.17

- Date: 2026-04-30
- Version: 1.2.17
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.2.17
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.2.17

- **added** — Add `respond` decision to HITL middleware

Changes since langchain==1.2.16

release(langchain): 1.2.17 (#37110)
feat(langchain): add `respond` decision to HITL middleware (#37095)

### 1.2.16

- Date: 2026-04-29
- Version: 1.2.16
- Original notes: https://github.com/langchain-ai/langchain/releases/tag/langchain%3D%3D1.2.16
- Permalink: https://whatsnew.fyi/product/langchain/releases/1.2.16

- **changed** — Stop inlining agent state into tool-dispatch Sends to improve performance
- **added** — Add content-block-centric streaming (v2) in core
- **fixed** — Honor max_retries parameter in Fireworks integration
- **added** — Add ls_agent_type tag on create_agent calls
- **changed** — Add LRU cache when resolving types for improved performance

Changes since langchain==1.2.15

release(langchain): 1.2.16 (#37085)
chore(docs): update x handle references (#37081)
release(anthropic): 1.4.2 (#37061)
perf(langchain): stop inlining agent state into tool-dispatch `Send`s (#36960)
docs(langchain): correct import paths in agent middleware docstrings (#36987)
hotfix: bump min core versions (#36996)
release(openai): 1.2.1 (#36995)
feat(core): add content-block-centric streaming (v2) (#36834)
fix(fireworks): honor `max_retries` (#36973)
release(openai): 1.2.0 (#36961)
docs(langchain): correct import path in `ModelCallLimitMiddleware` docstring (#36895)
feat(langchain): `ls_agent_type` tag on `create_agent` calls (#36774)
release(core): release 1.3.0 (#36851)
chore: bump langsmith from 0.6.3 to 0.7.31 in /libs/langchain_v1 (#36800)
chore(deps): bump pytest to `9.0.3` (#36801)
chore: bump pytest from 9.0.2 to 9.0.3 in /libs/langchain_v1 (#36717)
chore: bump pillow from 12.1.1 to 12.2.0 in /libs/langchain_v1 (#36710)
test(langchain): test cache hit/miss on lru (#36659)
perf(langchain): add lru cache when resolving types (#36649)
perf(langchain): add another init test with middleware (#36644)
chore(core): reduce streaming metadata / perf (#36588)
perf(langchain): add benchmark command (#36641)
chore: bump cryptography from 46.0.6 to 46.0.7 in /libs/langchain_v1 (#36619)
docs(langchain): add missing baseten and litellm to `init_chat_model` (#36562)
release(core): 1.2.26 (#36511)