langgenius/dify 1.9.0 on GitHub

🚀 Introduction

In Dify 1.9.0, we are introducing two major new capabilities: the Knowledge Pipeline and the Queue-based Graph Engine.

This is a beta release, and we hope to explore these improvements together with you and gather your feedback. The Knowledge Pipeline provides a modularized and extensible workflow for knowledge ingestion and processing, while the Queue-based Graph Engine makes workflow execution more robust and controllable. We believe these will help you build and debug AI applications more smoothly, and we look forward to your experiences to help us continuously improve.

📚 Knowledge Pipeline

✨ Introduction

With the brand-new orchestration interface for knowledge pipelines, we introduce a fundamental architectural upgrade that reshapes how document processing are designed and executed, providing a more modular and flexible workflow that enables users to orchestrate every stage of the pipeline. Enhanced with a wide range of powerful plugins available in the marketplace, it empowers users to flexibly integrate diverse data sources and processing tools. Ultimately, this architecture enables building highly customized, domain-specific RAG solutions that meet enterprises’ growing demands for scalability, adaptability, and precision.

❓ Why Do We Need It?

Previously, Dify's RAG users still encounter persistent challenges in real-world adoption — from inaccurate knowledge retrieval and information loss to limited data integration and extensibility. Common pain points include:

🔗 restricted integration of data sources
🖼️ missing critical elements such as tables and images
✂️ suboptimal chunking results

All of them lead to poor answer quality and hinder the model's overall performance.

In response, we reimagined RAG in Dify as an open and modular architecture, enabling developers, integrators, and domain experts to build document processing pipelines tailored to their specific requirements—from data ingestion to chunk storage and retrieval.

🛠️ Core Capabilities

🧩 Knowledge Pipeline Architecture

The Knowledge Pipeline is a visual, node-based orchestration system dedicated to document ingestion. It provides a customizable way to automate complex document processing, enabling fine-grained transformations and bridging raw content with structured, retrievable knowledge. Developers can build workflows step by step, like assembling puzzle pieces, making document handling easier to observe and adjust.

📑 Templates & Pipeline DSL

⚡ Start quickly with official templates
🔄 Customize and share pipelines by importing/exporting via DSL for easier reusability and collaboration

🔌 Customizable Data Sources & Tools

Each knowledge base can support multiple data sources. You can seamlessly integrate local files, online documents, cloud drives, and web crawlers through a plugin-based ingestion framework. Developers can extend the ecosystem with new data-source plugins, while marketplace processors handle specialized use cases like formulas, spreadsheets, and image parsing — ensuring accurate ingestion and structured representation.

🧾 New Chunking Strategies

In addition to General and Parent-Child modes, the new Q&A Processor plugin supports Q&A structures. This expands coverage for more use cases, balancing retrieval precision with contextual completeness.

🖼️ Image Extraction & Retrieval

Extract images from documents in multiple formats, store them as URLs in the knowledge base, and enable mixed text-image outputs to improve LLM-generated answers.

🧪 Test Run & Debugging Support

Before publishing a pipeline, you can:

▶️ Execute a single step or node independently
🔍 Inspect intermediate variables in detail
👀 Preview string variables as Markdown in the variable inspector

This provides safe iteration and debugging at every stage.

🔄 One-Click Migration from Legacy Knowledge Bases

Seamlessly convert existing knowledge bases into the Knowledge Pipeline architecture with a single action, ensuring smooth transition and backward compatibility.

🌟 Why It Matters

The Knowledge Pipeline makes knowledge management more transparent, debuggable, and extensible. It is not the endpoint, but a foundation for future enhancements such as multimodal retrieval, human-in-the-loop collaboration, and enterprise-level data governance. We’re excited to see how you apply it and share your feedback.

⚙️ Queue-based Graph Engine

❓ Why Do We Need It?

Previously, designing workflows with parallel branches often led to:

🌀 Difficulty managing branch states and reproducing errors
❌ Insufficient debugging information
🧱 Rigid execution logic lacking flexibility

These issues reduced the usability of complex workflows. To solve this, we redesigned the execution engine around queue scheduling, improving management of parallel tasks.

🛠️ Core Capabilities

📋 Queue Scheduling Model

All tasks enter a unified queue, where the scheduler manages dependencies and order. This reduces errors in parallel execution and makes topology more intuitive.

🎯 Flexible Execution Start Points

Execution can begin at any node, supporting partial runs, resumptions, and subgraph invocations.

🌊 Stream Processing Component

A new ResponseCoordinator handles streaming outputs from multiple nodes, such as token-by-token LLM generation or staged results from long-running tasks.

🕹️ Command Mechanism

With the CommandProcessor, workflows can be paused, resumed, or terminated during execution, enabling external control.

🧩 GraphEngineLayer

A new plugin layer that allows extending engine functionality without modifying core code. It can monitor states, send commands, and support custom monitoring.

Quickstart

Prerequisites
- Dify version: 1.9.0 or higher
How to Enable
- Enabled by default, no additional configuration required.
- Debug mode: set DEBUG=true to enable DebugLoggingLayer.
- Execution limits:
  - WORKFLOW_MAX_EXECUTION_STEPS=500
  - WORKFLOW_MAX_EXECUTION_TIME=1200
  - WORKFLOW_CALL_MAX_DEPTH=10
- Worker configuration (optional):
  - WORKFLOW_MIN_WORKERS=1
  - WORKFLOW_MAX_WORKERS=10
  - WORKFLOW_SCALE_UP_THRESHOLD=3
  - WORKFLOW_SCALE_DOWN_IDLE_TIME=30
- Applies to all workflows.

More Controllable Parallel Branches

Execution Flow:

Start ─→ Unified Task Queue ─→ WorkerPool Scheduling
                          ├─→ Branch-1 Execution
                          └─→ Branch-2 Execution
                                  ↓
                            Aggregator
                                  ↓
                                  End

Improvements:
1. All tasks enter a single queue, managed by the Dispatcher.
2. WorkerPool auto-scales based on load.
3. ResponseCoordinator manages streaming outputs, ensuring correct order.

Example: Command Mechanism

from core.workflow.graph_engine.manager import GraphEngineManager

# Send stop command
GraphEngineManager.send_stop_command(
    task_id="workflow_task_123",
    reason="Emergency stop: resource limit exceeded"
)

Note: pause/resume functionality will be supported in future versions.

Example: GraphEngineLayer

GraphEngineLayer Example

FAQ

Is this release focused on performance?
No. The focus is on stability, clarity, and correctness of parallel branches. Performance improvements are a secondary benefit.
What events can be subscribed to?
- Graph-level: GraphRunStartedEvent, GraphRunSucceededEvent, GraphRunFailedEvent, GraphRunAbortedEvent
- Node-level: NodeRunStartedEvent, NodeRunSucceededEvent, NodeRunFailedEvent, NodeRunRetryEvent
- Container nodes: IterationRunStartedEvent, IterationRunNextEvent, IterationRunSucceededEvent, LoopRunStartedEvent, LoopRunNextEvent, LoopRunSucceededEvent
- Streaming output: NodeRunStreamChunkEvent
How can I debug workflow execution?
- Enable DEBUG=true to view detailed logs.
- Use DebugLoggingLayer to record events.
- Add custom monitoring via GraphEngineLayer.

Future Plans

This beta release is just the beginning. Upcoming improvements include:

Debugging Tools: A visual interface to view execution states and variables in real time.
Intelligent Scheduling: Optimize scheduling strategies using historical data.
More Complete Command Support: Add Pause/Resume, breakpoint debugging.
Human in the Loop: Support human intervention during execution.
Subgraph Functionality: Enhance modularity and reusability.
Multimodal Embedding: Support richer content types beyond text.

We look forward to your feedback and experiences to make the engine more practical.

Upgrade Guide

Important

After upgrading, you must run the following migration to transform existing datasource credentials. This step is required to ensure compatibility with the new version:

uv run flask transform-datasource-credentials

Docker Compose Deployments

Back up your customized docker-compose YAML file (optional)

cd docker
cp docker-compose.yaml docker-compose.yaml.$(date +%s).bak

Get the latest code from the main branch

git checkout 1.9.0
git pull origin 1.9.0

Stop the service. Please execute in the docker directory

docker compose down

Back up data

tar -cvf volumes-$(date +%s).tgz volumes

Upgrade services

docker compose up -d

Migrate data after the container starts

docker exec -it docker-api-1 uv run flask transform-datasource-credentials

Source Code Deployments

Stop the API server, Worker, and Web frontend Server.
Get the latest code from the release branch:

git checkout 1.9.0

Update Python dependencies:

cd api
uv sync

Then, let's run the migration script:

uv run flask db upgrade
uv run flask transform-datasource-credentials

Important

If you are currently using version v2.0.0-beta.1 or v2.0.0-beta.2, you must run the following migration script to upgrade.

[WARNING] This is a destructive operation that will result in data loss.

This migration script will permanently delete the following data:

All existing Knowledge Pipelines
All related Dataset Credentials

Please ensure you have backed up any necessary information before proceeding.

uv run flask db downgrade cf7c38a32b2d
uv run flask db upgrade
uv run flask transform-datasource-credentials

Finally, run the API server, Worker, and Web frontend Server again.

What's Changed

fix: improve error logging for vector search operation in MyScale by @ZeroZ-lab in #25087
Fix jsonschema compliance: use number instead of float by @hyongtao-code in #25049
chore: Updated pnpm version to 10.15.1 by @17hz in #25065
chore: comply to RFC 6750 and improve bearer token split by @NeatGuyCoding in #24955
db internal server error by @xu15031483273 in #24947
chore: enhance basedpyright-check script to support path arguments by @laipz8200 in #25108
[Chore/Refactor] Improve type safety and resolve type checking issues by @laipz8200 in #25104
make icon consistent in dropdown by @jubinsoni in #25109
chore: improved trace info for generating conversation name by @IthacaDream in #25118
Chore: remove dead code in class Graph by @hyongtao-code in #22791
feat: add test containers based tests for workflow converter by @NeatGuyCoding in #25115
fix: align text color in dark mode for config var type selector by @qiaofenlin in #25121
fix: old custom model not display credential name by @hjlarry in #25112
fix: TypeSelector component style by @17hz in #25124
fix png jpeg export by @jubinsoni in #25110
Minor fix: correct get_app_model mode for delete() by @hyongtao-code in #25082
fix recommended apps reading from db logic by @TonloOO in #25071
chore: Bump weaviate-client to latest v3 version by @DavideDelbianco in #25096
fix: EndUser not bound to Session when plugin invokes callback by @IthacaDream in #25132
feat(workflow): Allow paste node into nested block by @CrabSAMA in #24234
feat: add TypeScript type safety for i18next with automated maintenance by @lyzno1 in #25152
fix: update account profile by @IthacaDream in #25150
fix: reset password by @IthacaDream in #25172
Remove unused mypy script by @hyongtao-code in #25177
[Chore/Refactor] Improve type checking configuration by @laipz8200 in #25185
fix: Ensure the order of execution steps are correct when logging with Weave by W&B by @xprilion in #25183
example regexp exec by @asukaminato0721 in #25200
fix: Remove invalid key from firecrawl request payload. by @IMOKURI in #25190
fix: chunk detail modal answer not wrap line by @lcedaw in #25203
to RefObject by @asukaminato0721 in #25192
Chore: clean some # type: ignore by @hyongtao-code in #25157
refactor: improve plugin version validation to support full semantic versioning by @laipz8200 in #25161
example of type button by @asukaminato0721 in #25224
example of readonly by @asukaminato0721 in #25220
fix: child chunk API 404 due to UUID type comparison by @kenwoodjw in #25234
fix: loop node doesn't exit when it react the condition #24717 by @coolfinish in #24844
example of lazy by @asukaminato0721 in #25216
example of remove useEffect by @asukaminato0721 in #25212
add type annotations for Python SDK ChatClient Class by @EchterTimo in #24018
feat: add reasoning format processing to LLMNode for tag handling by @ultramancode in #23313
Feature add test containers add document to index by @NeatGuyCoding in #25251
add more dataclass by @asukaminato0721 in #25039
remove bare list, dict, Sequence, None, Any by @asukaminato0721 in #25058
feat(web): add Progressive Web App (PWA) support by @laipz8200 in #25274
fix: standardize text color in install form to text-secondary by @laipz8200 in #25272
chore: translate i18n files and update type definitions by @github-actions[bot] in #25260
fix: resolve typing errors in configs module by @laipz8200 in #25268
Feature add test containers batch clean document by @NeatGuyCoding in #25287
example of no-unstable-context-value by @asukaminato0721 in #25279
fix: a failed index to be marked as created by @NeatGuyCoding in #25290
chore: ignore PWA generated files in version control by @laipz8200 in #25313
fix: remove workflow file preview docs by @lyzno1 in #25318
fix: update iteration node to use correct variable segment types by @laipz8200 in #25315
Fix: correct queryKey in useBatchUpdateDocMetadata and add test case by @hyongtao-code in #25327
fix: ensure consistent DSL export behavior across UI entry by @Kritoooo in #25317
[Chore/Refactor] Improve type annotations in models module by @laipz8200 in #25281
keep add and remove the same by @asukaminato0721 in #25277
Refactor：upgrade react19 ref as props by @zyileven in #25225
more assert by @asukaminato0721 in #24996
fix: 'curr_message_tokens' where it is not associated with a value #25307 by @qxo in #25308
chore: translate i18n files and update type definitions by @github-actions[bot] in #25334
Fix: Parameter Extractor Uses Correct Prompt for Prompt Mode in Chat Models by @ding113 in #24636
add more typing by @asukaminato0721 in #24949
chore: remove weird account login by @zalcit in #22247
feat: add test containers based tests for clean dataset task by @NeatGuyCoding in #25341
feat: add MCP server headers support #22718 by @Cluas in #24760
Minor fix: correct PrecessRule typo by @hyongtao-code in #25346
chore: translate i18n files and update type definitions by @github-actions[bot] in #25349
fix: ensure Performance Tracing button visible when no tracing provid… by @Kritoooo in #25351
feat: email register refactor by @zyssyz123 in #25344
Feature add test containers batch create segment to index by @NeatGuyCoding in #25306
fix: Incorrect URL Parameter Parsing Causes user_id Retrieval Error by @joeyhacker in #25261
feat: enable dsl export encrypt dataset id or not by @kenwoodjw in #25102
Revert "feat: email register refactor" by @zyssyz123 in #25367
refactor: update pyrightconfig.json to use ignore field for better type checking configuration by @laipz8200 in #25373
Fix/disable no constant binary expression by @MatriQ in #25311
Fix: use correct maxLength prop for verification code input by @hyongtao-code in #25371
fix: Fixed the X-axis scroll bar issue in the LLM node settings panel by @hudenghui in #25357
fix: tenant_id was not specific when retrieval end-user in plugin backwards invocation wraps by @Yeuoly in #25377
minor fix: some translation mismatch by @NeatGuyCoding in #25386
Support Anchor Scroll In The Output Node by @yinuu in #25364
example of uuid by @asukaminato0721 in #25380
feat: add test containers based tests for clean notion document task by @NeatGuyCoding in #25385
fix(container-test): batch create segment position sort by @Nov1c444 in #25394
Feat/credential policy by @GareArc in #25151
fix: credential not allow to use in load balancing by @zxhlyh in #25401
fix(mcp): current_user not being set in MCP requests by @Nov1c444 in #25393
Fix flask response: 200 -> {}, 200 by @hyongtao-code in #25404
feat: Add customized json schema validation by @WTW0313 in #25408
Fix:About the error problem of creating an empty knowledge base interface in service_api by @ttz12345 in #25398
feat: change history by supplementary node information by @AllesOderNicht in #25294
fix: avatar is not updated after setted by @iamjoel in #25414
Revert "example of remove useEffect" by @crazywoola in #25418
add typing to all wraps by @asukaminato0721 in #25405
Fix: judgement of open in explore by @JzoNgKVO in #25420
Fix: Use --fix flag instead of --fix-only in autofix workflow by @laipz8200 in #25425
Add lib test by @ZeroZ-lab in #25410
refactor: update pyrightconfig to scan all API files by @laipz8200 in #25429
fix(mcp): properly handle arrays containing both numbers and strings by @Nov1c444 in #25430
Chore: thanks to bump-pydantic by @hyongtao-code in #25437
Fix basedpyright type errors by @laipz8200 in #25435
fix: imporve the condition for stopping the think timer. by @17hz in #25365
add rule for strenum by @asukaminato0721 in #25445
Fix: Add Password Validation to Account Creation by @newtonjose in #25382
Feature add test containers create segment to index task by @NeatGuyCoding in #25450
fix: inner_api get_user_tenant by @IthacaDream in #25462
test: imporve by @ZeroZ-lab in #25461
refactor: Migrate part of the console basic API module to Flask-RESTX by @liugddx in #24732
update sql in batch by @asukaminato0721 in #24801
fix: Remove sticky positioning from workflow component fields by @WTW0313 in #25470
Fix 500 in dataset page. by @Eric-Guo in #25474
Feat/enteprise cd by @GareArc in #25485
Feat/enteprise cd by @GareArc in #25508
chore: support Zendesk widget by @nite-knite in #25517
security(api): fix privilege escalation vulnerability in model config and chat message APIs by @QuantumGhost in #25518
feat: add test containers based tests for deal dataset vector index by @NeatGuyCoding in #25545
fix: app name overflow by @JeeekXY in #25551
Fix log time display bug by @liqigang63 in #25475
test: remove print code by @ZeroZ-lab in #25481
Fix: correct has_more pagination logic in get_conversational_variable by @hyongtao-code in #25484
feat: add test containers based tests for delete segment from index task by @NeatGuyCoding in #25564
fix: delete get upload file endpoint by @WTW0313 in #25543
add autofix pnpm by @asukaminato0721 in #25557
Fixes #25530 by @zhuzeyu22 in #25531
chore(deps): bump mermaid from 11.4.1 to 11.10.0 in /web by @dependabot[bot] in #25521
Remove redundant parse_args call in WorkflowByIdApi.patch by @hyongtao-code in #25498
use autospec=True in mock by @asukaminato0721 in #25497
fix: single step system file error by @Nov1c444 in #25533
fix(date-picker): handle string date to avoid crash by @JarsirLiu in #25522
Fix/disable no unsafe optional chaining by @MatriQ in #25553
fix: Support passing the default app mode when creating an app by @Kurumi1997 in #25142
Force update search method to keyword_search by @ZombieBlue in #25464
Feat/email register refactor by @zyssyz123 in #25369
fix: login security issue frontend by @iamjoel in #25571
refactor: centralize default end user session ID constant by @Yeuoly in #25416
chore: translate i18n files and update type definitions by @github-actions[bot] in #25575
fix: filter temporary edges from workflow draft sync by @JQSevenMiao in #25442
refactor: update API routes and documentation for console endpoints by @liugddx in #25554
fix segment deletion race condition by @kenwoodjw in #24408
feat: add type integer to VarType by @jingfelix in #25500
fix close button cannot be clicked when the browser page is zoomed out by @chengjoey in #25584
fix: Cannot modify values when startNode has defaultValue by @17hz in #25595
chore: adopt StrEnum and auto() for some string-typed enums by @Kritoooo in #25129
feat(stress-test): add comprehensive stress testing suite using Locust by @laipz8200 in #25617
Feature add test containers disable segment from index task by @NeatGuyCoding in #25631
fix: Multiple UX improvements for GotoAnything command palette by @lyzno1 in #25637
[Test] speed up Hypothesis strategies to avoid too_slow by @hyongtao-code in #25623
feat: add circular scrolling to GotoAnything command menu by @lyzno1 in #25662
fix: allow empty values in Variable Inspector by @yoichiojima-2 in #25644
feat: add test containers based tests for disable segments from index task by @NeatGuyCoding in #25660
feat: enhance Makefile with code quality commands and default help by @laipz8200 in #25655
example for all by @asukaminato0721 in #25666
chore: translate i18n files and update type definitions by @github-actions[bot] in #25645
bugfix: The randomly generated email by Faker actually corresponded to an existing account in the test database, causing the test to fail. by @liugddx in #25646
improve type hints using typing.Literal and add type annotations by @EchterTimo in #25641
chore: bump pnpm version to v10.16.0 by @lyzno1 in #25640
optimize _merge_splits function by using enumerate instead of manual index tracking by @ChasePassion in #25680
[Chore/Refactor] use all to specify export member. by @ritoban23 in #25681
feat: user message support generate prompt by @iamjoel in #25689
refactor: update API routes and documentation for app and datasets endpoints by @liugddx in #25628
dependabot by @asukaminato0721 in #25677
chore(deps): bump transformers from 4.53.3 to 4.56.1 in /api by @dependabot[bot] in #25696
chore(deps): bump clickzetta-connector-python from 0.8.102 to 0.8.104 in /api by @dependabot[bot] in #25697
chore(deps-dev): bump sass from 1.89.2 to 1.92.1 in /web by @dependabot[bot] in #25698
chore(deps): bump scheduler from 0.23.2 to 0.26.0 in /web by @dependabot[bot] in #25699
chore(deps-dev): bump @next/bundle-analyzer from 15.5.0 to 15.5.3 in /web by @dependabot[bot] in #25704
chore(deps): bump @lexical/text from 0.30.0 to 0.35.0 in /web by @dependabot[bot] in #25705
chore: add ast-grep rule to convert Optional[T] to T | None by @laipz8200 in #25560
fix: resolve devalue prototype pollution vulnerability by @laipz8200 in #25709
Release/e-1.8.1 by @GarfieldDai in #25613
fix(mcp): prevent masked headers from overwriting real values by @quicksandznzn in #25722
fix(web): custom-tool output_schema.properties missing type by @chengjoey in #25731
correct typos . by @zhanluxianshen in #25717
chore: translate i18n files and update type definitions by @github-actions[bot] in #25713
fix(tests): resolve order dependency in disable_segments_from_index_task tests by @laipz8200 in #25737
Bugfix: When i change the loop variable, 'Loop Termination Condition' wi… by @minhuaF in #25695
Feature: add test containers document indexing task by @NeatGuyCoding in #25684
chat remove transparency from chat bubble in dark mode by @jubinsoni in #24921
ruff check preview by @asukaminato0721 in #25653
Enhance LLM model configuration validation to include active status c… by @GareArc in #25759
refactor: remove Claude-specific references from documentation files by @laipz8200 in #25760
Fix/lindorm vdb optimize by @AlwaysBluer in #25748
Fix incorrect API endpoint routing from PR #25628 by @laipz8200 in #25778
fix: remove billing cache when add or delete app or member by @zyssyz123 in #25885
feat: knowledge pipeline by @laipz8200 in #25360
fix: handle None description in MCP tool transformation by @Nov1c444 in #25872
feat(api/commands): add migrate-oss to migrate from Local/OpenDAL to … by @quicksandznzn in #25828
chore: improve opendal storage and ensure closing file after reading files in load_stream method by @bowenliang123 in #25874
chore: add missing template translations in ja-JP by @lyzno1 in #25892
fix: update Python base image to use bullseye variant by @ZeroZ-lab in #25895
fix: datasource pinned list by @zxhlyh in #25896
fix: workflow logs list by @zxhlyh in #25903
chore(workflows): remove redundant eslint command from style workflow by @17hz in #25900
perf(web): optimize ESLint performance with concurrency flag and remove oxlint by @17hz in #25899
chore: Update the value of sys.dialogue_count to start from 1. by @Nov1c444 in #25905
refactor: simplify portal interactions and manage state in Configure component by @WTW0313 in #25906
feat: make SQLALCHEMY_POOL_TIMEOUT configurable by @saurabhy27 in #25468
fix: remote code execution in email endpoints by @wylswz in #25753
fix: use invalid last run by @zxhlyh in #25911
fix(api): fix single stepping variable loading by @QuantumGhost in #25908
fix: workflow header style by @zxhlyh in #25922
fix: Message => str by @crazywoola in #25876
fix(datasets): retrieval_model null issue when updating dataset info by @quicksandznzn in #25907
fix: rollback aliyun_trace icon by @hieheihei in #25921
refactor: replace useInvalid with useInvalidCustomizedTemplateList by @WTW0313 in #25924
fix(api): enhance data handling in RagPipelineDslService to filter credentials by @Mairuis in #25926
refactor: replace print statements with proper logging by @laipz8200 in #25773
fix(makefile): correct uv project path for lint target by @laipz8200 in #25818
Fix: replace stdout prints with debug logging by @laipz8200 in #25931
chore: update post_create_command.sh to use dynamic workspace root for aliases by @ZeroZ-lab in #25913
Fix: enforce editor-only access to chat message logs by @laipz8200 in #25936
Chore: correct inconsistent logging and typo by @hyongtao-code in #25945
style: update GotoAnything component styling by @ZeroZ-lab in #25929
fix: refresh datasource list after install datasource by @zxhlyh in #25949
Refactor WorkflowService to handle missing default credentials gracef… by @GareArc in #25960
fix: update details display to conditionally show creator information by @WTW0313 in #25952
fix: enhance model parameter handling with advanced mode support and localization updates by @WTW0313 in #25963
fix: undefined match the wrong output schema by @iamjoel in #25971
Run import-linter within make lint by @laipz8200 in #25933
chore: translate i18n files and update type definitions by @github-actions[bot] in #25964
fix: ensure original response are maintained by yielding text messages in ApiTool (#23456) by @Yeuoly in #25973
feat(goto-anything): add RAG pipeline node search by @ZeroZ-lab in #25948
fix: resolve chat sidebar UI bugs for hover panel and dropdown menu by @lyzno1 in #25813
Chore: remove dead code in datasource.utils by @hyongtao-code in #25984
fix: Statistics, like workflows, do not include debug data. by @jiangbo721 in #25979
chore(deps): bump authlib from 1.3.1 to 1.6.4 in /api by @dependabot[bot] in #26015
chore(deps): bump negotiator from 0.6.4 to 1.0.0 in /web by @dependabot[bot] in #26012
feat:support selecting different ftparser for OceanBase. by @longbingljw in #25970
Fix: use data.type instead of type when checking datasource node by @hyongtao-code in #25965
chore(deps-dev): bump eslint-plugin-storybook from 0.11.6 to 9.0.7 in /web by @dependabot[bot] in #26011
Add Full-Text & Hybrid Search Support to Baidu Vector DB and Update SDK, Closes #25982 by @WayneCao in #25983
fix: pass operation name to onUpdate callback in StatusItem component by @WTW0313 in #26019
chore(deps): bump abcjs from 6.5.1 to 6.5.2 in /web by @dependabot[bot] in #26018
Fix: correct regex for file-preview URL re-signing by @hyongtao-code in #25620
Align dev workflow branch triggers by @laipz8200 in #26029
fix: workflow by @zyssyz123 in #26030
chore(api): upgade Gevent to 25.9.1 by @QuantumGhost in #26026
chore(deps): bump boto3-stubs from 1.40.29 to 1.40.35 in /api by @dependabot[bot] in #26014
fix: comment out unused segmentation rule properties in RuleDetail component by @WTW0313 in #26031
fix: Correctly map source_url to preview_url in file fields by @lei-jia-xing in #25957
chore: remove mistakenly added trash file by @Stream29 in #26041
chore(api): adjust monkey patching in gunicorn.conf.py by @QuantumGhost in #26056
fix: iteration and loop node single step run by @Nov1c444 in #26036
chore(api): bump version by @QuantumGhost in #25917
fix: Ensure compatibility with old provider name when updating model credentials by @hjlarry in #26017
fix preview url by @JohnJyong in #26059

New Contributors

@xu15031483273 made their first contribution in #24947
@CrabSAMA made their first contribution in #24234
@xprilion made their first contribution in #25183
@IMOKURI made their first contribution in #25190
@coolfinish made their first contribution in #24844
@EchterTimo made their first contribution in #24018
@Kritoooo made their first contribution in #25317
@qxo made their first contribution in #25308
@ding113 made their first contribution in #24636
@Cluas made their first contribution in #24760
@joeyhacker made their first contribution in #25261
@yinuu made their first contribution in #25364
@ttz12345 made their first contribution in #25398
@AllesOderNicht made their first contribution in #25294
@newtonjose made their first contribution in #25382
@JeeekXY made their first contribution in #25551
@liqigang63 made their first contribution in #25475
@JarsirLiu made their first contribution in #25522
@Kurumi1997 made their first contribution in #25142
@JQSevenMiao made their first contribution in #25442
@chengjoey made their first contribution in #25584
@yoichiojima-2 made their first contribution in #25644
@ChasePassion made their first contribution in #25680
@ritoban23 made their first contribution in #25681
@minhuaF made their first contribution in #25695
@saurabhy27 made their first contribution in #25468
@longbingljw made their first contribution in #25970
@lei-jia-xing made their first contribution in #25957

Full Changelog: 1.8.1...1.9.0

langgenius/dify 1.9.0 1.9.0 – Orchestrating Knowledge, Powering Workflows on GitHub

🚀 Introduction

📚 Knowledge Pipeline

✨ Introduction

❓ Why Do We Need It?

🛠️ Core Capabilities

🧩 Knowledge Pipeline Architecture

📑 Templates & Pipeline DSL

🔌 Customizable Data Sources & Tools

🧾 New Chunking Strategies

🖼️ Image Extraction & Retrieval

🧪 Test Run & Debugging Support

🔄 One-Click Migration from Legacy Knowledge Bases

🌟 Why It Matters

⚙️ Queue-based Graph Engine

❓ Why Do We Need It?

🛠️ Core Capabilities

📋 Queue Scheduling Model

🎯 Flexible Execution Start Points

🌊 Stream Processing Component

🕹️ Command Mechanism

🧩 GraphEngineLayer

Quickstart

More Controllable Parallel Branches

Example: Command Mechanism

Example: GraphEngineLayer

FAQ

Future Plans

Upgrade Guide

Docker Compose Deployments

Source Code Deployments

What's Changed

New Contributors

langgenius/dify 1.9.0
1.9.0 – Orchestrating Knowledge, Powering Workflows

on GitHub