46 Pull requests merged by 24 people

• fix: add missing package xinference_client to pass vdb CI tests

  #13865 merged Feb 17, 2025

• Retrieval service optimization

  #13849 merged Feb 17, 2025

• fix: undefined attribute 'query' on MessageAnnotation

  #13852 merged Feb 17, 2025

• fix: markdown merge error

  #13853 merged Feb 17, 2025

• Introduce Plugins

  #13836 merged Feb 17, 2025

• Retrieval Service efficiency optimization

  #13543 merged Feb 17, 2025

• fix: ignore invalid marketplace sepc

  #13803 merged Feb 17, 2025

• Plugin/merge main 0217

  #13800 merged Feb 17, 2025

• fix: update trial expire time to 3/17

  #13796 merged Feb 17, 2025

• Revert "Feat: compliance report download"

  #13799 merged Feb 17, 2025

• Revert "Feat/compliance"

  #13798 merged Feb 17, 2025

• Feat: compliance report download

  #13282 merged Feb 17, 2025

• Feat/compliance

  #13548 merged Feb 17, 2025

• fix: add install a package to improve the accuracy of guessing mime type and file extension (main)

  #13752 merged Feb 16, 2025

• perf(message): optimize message loading and reduce SQL queries

  #13720 merged Feb 15, 2025

• fix: reranking_enable setting failed #13668

  #13721 merged Feb 14, 2025

• Non-Streaming Models Do Not Return Results Properly in _handle_invoke_result

  #13571 merged Feb 14, 2025

• fix:Knowledge Base with Parent-Child segment mode not support in Agent

  #13663 merged Feb 14, 2025

• Chore/format code

  #13691 merged Feb 14, 2025

• Fix/import old dsl

  #13718 merged Feb 14, 2025

• Fix: model icon position

  #13707 merged Feb 14, 2025

• perf: Implemented short-circuit evaluation for logical conditions

  #13674 merged Feb 13, 2025

• fix: error in obtaining end_to_node_id during conditional parallel execution

  #13673 merged Feb 13, 2025

• chore: update the agent model trigger

  #13639 merged Feb 13, 2025

• fix: add plugin_id to GenericProviderID format checking error

  #13679 merged Feb 13, 2025

• chore:Remove duplicate code, lines 8 to 27, same as lines 29 & 45 to 62.

  #13659 merged Feb 13, 2025

• fix: fix tongyi models blocking mode with incremental_output=stream

  #13620 merged Feb 13, 2025

• fix: agent thought not saved

  #13631 merged Feb 13, 2025

• fix: unexpected None embedding model provdier

  #13610 merged Feb 12, 2025

• fix document extractor node incorrectly processing doc and ppt files

  #12902 merged Feb 12, 2025

• feat: add Dify-Hook-Url to endpoint headers

  #13602 merged Feb 12, 2025

• fix: convert provider_id to plugin_provider_id in get_configurations

  #13596 merged Feb 12, 2025

• fix: refresh after install plugin

  #13593 merged Feb 12, 2025

• fix: add langgenius to list tool api

  #13578 merged Feb 12, 2025

• fix: no longer using old app detail cover when switch pathname

  #13585 merged Feb 12, 2025

• (1.0) fix: invalid default model provider

  #13572 merged Feb 12, 2025

• feat(vdb): add HNSW vector index for TiDB vector store with TiFlash

  #12043 merged Feb 12, 2025

• Feat: tool setting support variable

  #13465 merged Feb 12, 2025

• fix: changed topics/keywords to topic/keywords

  #13544 merged Feb 12, 2025

• chore:Remove useless brackets and format code

  #13479 merged Feb 11, 2025

• Ensure the 'inputs' field in /chat-messages takes effect every time

  #7955 merged Feb 11, 2025

• chore: add debug doc link

  #13537 merged Feb 11, 2025

• fix: iteration node log time error

  #13511 merged Feb 11, 2025

• fix: retry log running error

  #13472 merged Feb 11, 2025

• fix: installed plugin not show upgrade

  #13523 merged Feb 11, 2025

• fix: issue #13483 and #13434

  #13518 merged Feb 11, 2025

26 Pull requests opened by 25 people

• OpenVINO Model Server integration for local LLM inferencing.

  #13536 opened Feb 11, 2025

• fix: can not connect host.docker.internal host

  #13564 opened Feb 12, 2025

• fix the error of unable to retrieve url from file

  #13603 opened Feb 12, 2025

• build(deps): bump duckduckgo_search from 6.3.x to 7.3.2

  #13613 opened Feb 12, 2025

• fix: MessageLogModal closed unexpectedly

  #13617 opened Feb 12, 2025

• feat: Variable click jumps to source node

  #13623 opened Feb 12, 2025

• fix(docker-compose.yaml): Fix the issue where Docker Compose doesn't …

  #13637 opened Feb 13, 2025

• feat: When using redis cluster for caching, add global_prefix to all keys cached by redis

  #13652 opened Feb 13, 2025

• Fix: support grounding for Vertex AI

  #13661 opened Feb 13, 2025

• upgrade Unstructured version

  #13667 opened Feb 13, 2025

• Add new models for built-in Perplexity Tool

  #13694 opened Feb 13, 2025

• fix: github tool NoneType has no len()

  #13705 opened Feb 14, 2025

• feat: Add page_number attribute to document segments and update related retrieval logic

  #13742 opened Feb 14, 2025

• fix: use correct SlideSpeak API endpoint to verify API Key and fix typo

  #13756 opened Feb 14, 2025

• fix: replace old-style <br> tags to fix Mermaid rendering issues

  #13792 opened Feb 17, 2025

• chore: translate i18n files

  #13795 opened Feb 17, 2025

• feat:support clear all annotation

  #13797 opened Feb 17, 2025

• fix: thought time display error

  #13807 opened Feb 17, 2025

• set value of basePath

  #13810 opened Feb 17, 2025

• chore: prompt_message is actually assistant_message which is a bit am…

  #13839 opened Feb 17, 2025

• fix #13573

  #13843 opened Feb 17, 2025

• chore(build): bump poetry from 2.0 to 2.1

  #13851 opened Feb 17, 2025

• fix: "today" statistics are not displayed in chart view

  #13854 opened Feb 17, 2025

• Feat: support A/B (seamless) document updates for Knowledge Base

  #13855 opened Feb 17, 2025

• fix: change the config name and fix typo in description of the number of retrieval executors

  #13856 opened Feb 17, 2025

• Feat: Add pg_bigm for keyword search in pgvector

  #13876 opened Feb 18, 2025

176 Issues closed by 22 people

• the **agent-Monitoring-Public URL** field automatically populates an address without the port

  #13877 closed Feb 18, 2025

• Operational error when using managed PostgreSQL caused by too many concurrent connections

  #13875 closed Feb 18, 2025

• Images langgenius/dify-api:1.0.0 and langgenius/dify-web:1.0.0 are missing

  #13874 closed Feb 18, 2025

• 0.15.3版本无法从docker拉取镜像

  #13872 closed Feb 18, 2025

• After upgrading dify, the knowledge base and application data on the web page are gone.

  #13870 closed Feb 17, 2025

• docker compose upgrade can't pull

  #13868 closed Feb 17, 2025

• Add user input in iteration

  #13871 closed Feb 17, 2025

• Failures in VDB CI tests for missing xinference_client package

  #13867 closed Feb 17, 2025

• API interface for viewing file information

  #13863 closed Feb 17, 2025

• build error with docker

  #13847 closed Feb 17, 2025

• Dify workflow Integrated Search Engine, LLM node Unable to Retrieve the Output Variable“json" from the Search Engine tool.

  #13844 closed Feb 17, 2025

• Where is the Endpoint set

  #13840 closed Feb 17, 2025

• PluginDaemonBadRequestError：plugin verification has been enabled, and the plugin you want to install has a bad signature

  #13842 closed Feb 17, 2025

• Failed indexing csv files

  #13830 closed Feb 17, 2025

• Are Conversation Variables belong to Short-term memory or Long-term memory

  #13816 closed Feb 17, 2025

• The length of output variable result must be less than 80000 characters

  #13833 closed Feb 17, 2025

• How to recall query-relevant memory from "Conversation Variables "

  #13821 closed Feb 17, 2025

• Application Not Available 404

  #13826 closed Feb 17, 2025

• The file upload to the knowledge base is always stuck in the queue.

  #13829 closed Feb 17, 2025

• embedding 太慢，我能否通过增加worker的实例来加速文档的索引

  #13818 closed Feb 17, 2025

• How to config mail in .env.example pls help me to give a example thanks very much

  #13825 closed Feb 17, 2025

• Reached maximum retries (3) for URL /files/0f34580d-cbb1-4f0e-b11f-5c024de12acf/......

  #13786 closed Feb 17, 2025

• how to view log in code node in workflow

  #13827 closed Feb 17, 2025

• failed on attempt 4: Request URL is missing an 'http://' or 'https://' protocol

  #13814 closed Feb 17, 2025

• 请问通过API 访问/chat-messages 接口如何增加上下文知识库和前置一些数据信息RAG

  #13824 closed Feb 17, 2025

• The error above is from ssrf-proxy.py:62, which only triggers when uploading a URL file for chat on the frontend page of Dify

  #13828 closed Feb 17, 2025

• docker failed to run dify

  #13820 closed Feb 17, 2025

• Knowledge base retrieval efficiency problem

  #13553 closed Feb 17, 2025

• How to configure text-to-speech and voice-to-text models in the functional options of the local deployment workflow interface

  #13812 closed Feb 17, 2025

• Local deployment A problem occurs after the nginx mapping port is changed to 8080

  #13811 closed Feb 17, 2025

• Remove the `docker-legacy` directory

  #13793 closed Feb 17, 2025

• Source code private deployment exception

  #13806 closed Feb 17, 2025

• How to deploy with subdirectory?

  #13805 closed Feb 17, 2025

• use text-embeddings-inference get "413 Payload Too Large"

  #12263 closed Feb 17, 2025

• comfyUI workflow node outputs a fixed image

  #12288 closed Feb 17, 2025

• Summarise Entire Document

  #12225 closed Feb 17, 2025

• The node of http request support streaming request

  #11830 closed Feb 17, 2025

• Can Gemini add the function of Grounding with Google Search?

  #13790 closed Feb 17, 2025

• HacKear instagram Gratis con una web, Como recuperar cuentas de insta en 2025

  #13785 closed Feb 16, 2025

• Issue: mimetypes Module Fails to Detect MIME Types in Minimal Docker Images

  #13717 closed Feb 16, 2025

• docker compose up!!!!!

  #13782 closed Feb 16, 2025

• knowledge-retrieval: rerank model can’t be closed

  #13770 closed Feb 16, 2025

• The output of the DeepSeek R1 model on Azure AI Studio in the "Debug and Preview" interface does not display the thinking process interface correctly.

  #13776 closed Feb 16, 2025

• Chinese characters in the workflow are converted to Unicode.

  #13778 closed Feb 16, 2025

• Connection reset by peer when setting up SearXNG extention

  #13775 closed Feb 16, 2025

• Max file upload issue

  #13774 closed Feb 16, 2025

• LLM Node cannot read file with a custom field type of "File" when using a multimodal model

  #12269 closed Feb 15, 2025

• Missing Parameter Types in Quick Tool Integration Documentation

  #12256 closed Feb 15, 2025

• After multiple invocations of the same tool, the returned result is overwritten by the last invocation.

  #12260 closed Feb 15, 2025

• Regarding the issue of email notifications for joining a workspace in local deployment.

  #13771 closed Feb 15, 2025

• i use dify api and set "response_mode": "streaming" but the EventStream show all the message at same time

  #13699 closed Feb 15, 2025

• (reopen issue) How to configure so that dify can be accessed through nginx domain name + path (for example: https://my.com/dify)

  #13764 closed Feb 15, 2025

• Regarding the issue that the JSON in the HTTP tool cannot accept objects. In the current version, JSON can only accept strings. However, there will be escape characters in the JSON string processed by the code, and the HTTP request cannot be made normally.

  #13765 closed Feb 15, 2025

• Spam wechat

  #13766 closed Feb 15, 2025

• Unable to add models like Gemini, have already tried setting SSRF_PROXY_HTTP_URL

  #13762 closed Feb 15, 2025

• perf: Optimize chat message loading performance and message ordering

  #13723 closed Feb 15, 2025

• Too many incorrect password attempt

  #13763 closed Feb 15, 2025

• Still Happening on CLOUD hosted: Application error: a client-side exception has occurred (see the browser console for more information)

  #13757 closed Feb 15, 2025

• Why must sys.query be used compulsorily in the USER of the LLM node in the chatflow?

  #13758 closed Feb 15, 2025

• Unable to display picture

  #13749 closed Feb 14, 2025

• Collapse the chat session column.

  #12235 closed Feb 14, 2025

• Vision processing continues to work when disabled in LLM node

  #12220 closed Feb 14, 2025

• Weaviate supports Chinese bm25

  #12223 closed Feb 14, 2025

• The agree function of the chat dialog answer is not available

  #12212 closed Feb 14, 2025

• Common llm response processing

  #12211 closed Feb 14, 2025

• Add Japanese Text keyword extractor usch as MeCab/Janome

  #12204 closed Feb 14, 2025

• The Q&A mode always fails.

  #12142 closed Feb 14, 2025

• got a Application error: a client-side exception has occurred (see the browser console for more information).

  #13751 closed Feb 14, 2025

• The custom tool for multipart/form-data requests encountered a 'Missing boundary in multipart' error

  #13684 closed Feb 14, 2025

• Does Dify allow each user to have their own personal knowledge base?

  #13744 closed Feb 14, 2025

• Not support for Azure OpenAI o3-mini

  #13745 closed Feb 14, 2025

• can't close rerank model used

  #13668 closed Feb 14, 2025

• Meta endpoint! Unexpected status code: 502, with response body: None.

  #13722 closed Feb 14, 2025

• Failed to set up HTTPS

  #13741 closed Feb 14, 2025

• 如何把大模型回答内容保存成txt到本地？

  #13737 closed Feb 14, 2025

• 如何配置才能使得通过nginx的域名+path（比如：https://my.com/dify）进行代理访问到dify

  #13735 closed Feb 14, 2025

• Non-Streaming Models Do Not Return Results Properly in _handle_invoke_result

  #13569 closed Feb 14, 2025

• document extractor takes too long to parse or fails to parse PDFs

  #12516 closed Feb 14, 2025

• In the old process, an optional variable input was added, and using this variable for branch judgment caused an error in the previous session.

  #13730 closed Feb 14, 2025

• DeepSeek-R1 bot with VolcEngine API Works in Cherry Studio but Not in Dify

  #13729 closed Feb 14, 2025

• The latest version of dify. Does the chat messages API still support the parent_cessage_id parameter. Based on the already replied answer, regenerate the answer

  #13727 closed Feb 14, 2025

• docker compose has an erros in api container:SyntaxWarning: invalid escape sequence '\_'

  #13731 closed Feb 14, 2025

• Agent unable to use parent-child retrieval knowledge base

  #12645 closed Feb 14, 2025

• Knowledge Base with Parent-Child segment mode not support in Agent

  #13662 closed Feb 14, 2025

• 节点输出变量无法新增和更改变量类型

  #13712 closed Feb 14, 2025

• Playwright in Code Block

  #13696 closed Feb 14, 2025

• The HTTP component of the workflow reports a 403 error when calling the remote service API.

  #13702 closed Feb 14, 2025

• Multi-Agent System Support by Integrating Agents within Workflows

  #13704 closed Feb 14, 2025

• Storage Requirements for Multimedia File Generation

  #13695 closed Feb 14, 2025

• Published app webapp link timing out (took too long to respond)

  #13501 closed Feb 13, 2025

• After I updated to the latest version, webpages sneaked into other applications cannot be accessed.

  #13693 closed Feb 13, 2025

• Application error: a client-side exception has occurred(see the browser console for more information)

  #13690 closed Feb 13, 2025

• Knowledge Retrieval returns different results in Chat/Workflow and in Retrieval Testing

  #13622 closed Feb 13, 2025

• Optimize condition evaluation to exit early based on operator in group results processing

  #13686 closed Feb 13, 2025

• After dify was updated to 0.15.3, kimi's moonshot-v1-8k was using an error. api untunable

  #13685 closed Feb 13, 2025

• LLM Stuck at Question Classifier Node

  #13628 closed Feb 13, 2025

• Error when using Kimi model

  #13681 closed Feb 13, 2025

• Models imported in AWS Bedrock

  #13675 closed Feb 13, 2025

• ChatFlow Fails to Retrieve doc Parameter in Subsequent API Requests When Initially Omitted

  #13676 closed Feb 13, 2025

• deployed in a local Docker environment, but encountered a 404 error when accessing http://localhost/install

  #13671 closed Feb 13, 2025

• Why is LDAP login not supported

  #13678 closed Feb 13, 2025

• Application error: a client-side exception has occurred (see the browser console for more information)

  #13677 closed Feb 13, 2025

• full text search not work in chinese character in 0.15.3

  #13643 closed Feb 13, 2025

• The image format of the interface message is incorrect

  #13658 closed Feb 13, 2025

• During multiple chatflow conversations, the title is not automatically extracted and appears as New conversation, version 0.15.3

  #13653 closed Feb 13, 2025

• deepseek r1 model dialog format questions for local or cloud services

  #13656 closed Feb 13, 2025

• dify0.15.3 Why is there no thought process when using ds-r1 of Alibaba Cloud

  #13645 closed Feb 13, 2025

• Cannot open WebApp in WeCom workspace app, on MacOS

  #13644 closed Feb 13, 2025

• API and Worker containers are unable to connect to Redis

  #13641 closed Feb 13, 2025

• how to use dify with cline

  #13635 closed Feb 13, 2025

• chatflow工作的HTTP节点无法上传文件

  #13634 closed Feb 13, 2025

• How to add a child chunk to a parent chunk via API?

  #13629 closed Feb 13, 2025

• IF ELSE can cause the Stream output of LLM nodes to fail. again

  #13614 closed Feb 12, 2025

• When using OpenAI-API-compatible, the <think> tag is not returned

  #13621 closed Feb 12, 2025

• When I use Prompt Generator, I got a error: Failed to generate rule config. Error: [openai] Error: 1 validation error for LLMResult\nmodel\n Input should be a valid string

  #12177 closed Feb 12, 2025

• Can Dify automatically distinguish between think and reply content? Dify是否可以自动区分think和reply内容？

  #13611 closed Feb 12, 2025

• Will there be any problems replacing the <think></think> blocks in DeepSeek-R1 series output with quote markers to optimize the display effect?

  #13409 closed Feb 12, 2025

• Jina Reader API issue

  #13608 closed Feb 12, 2025

• Files with HTTP nodes exceeding 1M in the workflow cannot be uploaded

  #13609 closed Feb 12, 2025

• http nodes do not support custom response structs, only body strings

  #13607 closed Feb 12, 2025

• I can't receive the URL of the file after uploading it

  #13606 closed Feb 12, 2025

• Document extractor node incorrectly handles doc and ppt files

  #12889 closed Feb 12, 2025

• Application error: a client-side exception has occurred(see the browser console for more information)

  #13597 closed Feb 12, 2025

• full text search failed in 0.12.1 but work in 0.15.1

  #13598 closed Feb 12, 2025

• Firecrawl extraction doesn't apply "only_main_content" parameter when set to true in v0.15.2

  #13019 closed Feb 12, 2025

• stopped by user，not the same issue as 8874

  #13589 closed Feb 12, 2025

• When chatflow api /chat-messages passes in conversation_id (i.e. multiple rounds of conversation), inputs cannot be updated

  #13588 closed Feb 12, 2025

• Overview page card switch error display

  #13587 closed Feb 12, 2025

• Suggestions for Enhancements to Support Long Models and System Prompts

  #13584 closed Feb 12, 2025

• agent nodes cannot be added to the dify Community Edition workflow

  #13582 closed Feb 12, 2025

• Document vectorization failed

  #13583 closed Feb 12, 2025

• Unable to automatically perform embedding processing after uploading documents using the API?

  #13579 closed Feb 12, 2025

• Can you update to the latest image version, 0.15.4?

  #13580 closed Feb 12, 2025

• [Agent] ReAct mode cannot display echarts/mermaid correctly

  #13575 closed Feb 12, 2025

• Can not save keywords when editing knowledge base on windows.

  #12402 closed Feb 12, 2025

• API Returns Empty Data for Conversations and Messages Endpoints

  #12413 closed Feb 12, 2025

• How to disable to output that used tools to in chat with agent.

  #12412 closed Feb 12, 2025

• process_rule mode does not contemplate the hierarchical option

  #12506 closed Feb 12, 2025

• Creating Knowledge Base Error

  #12532 closed Feb 12, 2025

• Error Occurs When Setting DIFY_PORT to a Value Other Than 5001

  #13568 closed Feb 12, 2025

• feat: add HNSW vector index for TiDB vector store

  #12332 closed Feb 12, 2025

• conversaion not found

  #11753 closed Feb 12, 2025

• conversation Not Exists

  #13567 closed Feb 12, 2025

• Document stuck at Queuing

  #13194 closed Feb 12, 2025

• Messages are recalled after errors

  #13344 closed Feb 12, 2025

• Add PandasAI

  #12324 closed Feb 12, 2025

• Support multi-agent

  #11882 closed Feb 12, 2025

• Is there a Python SDK or API available that would allow us to programmatically create and manage workflow nodes in Dify?

  #12067 closed Feb 12, 2025

• Can't use stream function calling with a vllm hosted qwen2.5-7B model

  #12122 closed Feb 12, 2025

• Allow modifying the base url of Groq

  #12185 closed Feb 12, 2025

• stopped by user error

  #13556 closed Feb 12, 2025

• The response status code of the large model is abnormal, indicating that the request concurrency is too high, causing the retry function to fail.

  #13557 closed Feb 12, 2025

• Cannot upload txt or csv files. Seems only image file supported?

  #13558 closed Feb 12, 2025

• Credentials validation failed with status code 403

  #13560 closed Feb 12, 2025

• How to perform indexing after uploading a file to the knowledge base through an API?

  #13565 closed Feb 12, 2025

• How to implement streaming response when a Dify agent invokes a workflow?

  #13551 closed Feb 12, 2025

• Knowledges metadata

  #13502 closed Feb 12, 2025

• The PC browser is fine, but when you use your mobile phone to access the chat interface, you get a white screen. This bug is very urgent and I hope it will be fixed soon. Thanks

  #13545 closed Feb 12, 2025

• Possible issue with # of words not being recalculated after retry

  #12174 closed Feb 11, 2025

• Top K behavior is incompatible between index method `Economical` and `High Quality`

  #12144 closed Feb 11, 2025

• Restore and continue the conversation in historical dialogues(logs)

  #12141 closed Feb 11, 2025

• The multi-branch workflow cannot be fully scheduled and executed.

  #13532 closed Feb 11, 2025

• Unify Function Call and Tool Call in Agent Mode

  #13534 closed Feb 11, 2025

• An error message was displayed when an agent body was created in dify using ollama's local model API 400

  #13538 closed Feb 11, 2025

• running dify using docker compose on macbook fails

  #13540 closed Feb 11, 2025

• stuck at Queuing status when upload document always

  #13541 closed Feb 11, 2025

• There is an error when adding the DeepSeek appkey. 分享

  #13535 closed Feb 11, 2025

• Not Found: failed to connect to the endpoint: https://api.dify.ai/v1/retrieval

  #13531 closed Feb 11, 2025

• If you have forgotten your login credentials, what should you do to log in?

  #13528 closed Feb 11, 2025

• cannot access local variable 'result_text' where it is not associated with a value when using Deepseek R1 with nvidia api

  #13492 closed Feb 11, 2025

• In the iteration nodes within a workflow, if a node fails and retries, there will be a retry node that remains in a running state.

  #13386 closed Feb 11, 2025

• [moonshot] Error: 'str' object has no attribute 'name'

  #13524 closed Feb 11, 2025

• How to configure CELERY_BROKER_URL using redis-cluster?

  #13525 closed Feb 11, 2025

• [moonshot] Error: 'str' object has no attribute 'name'

  #13483 closed Feb 11, 2025

• Setup deepseek fail. "Credentials validation failed with status code 402"

  #13515 closed Feb 11, 2025

• Pycharm debugger gets stuck

  #12203 closed Feb 11, 2025

82 Issues opened by 73 people

• Add OpenAI-API-compatible Rerank error 400 Client Error

  #13879 opened Feb 18, 2025

• request to support SGLang llm, which comes from deepseek deployment

  #13878 opened Feb 18, 2025

• Feat: Add support for pgvector with Japanese keyword search

  #13873 opened Feb 17, 2025

• Help! Problem about chinese property description in Parameter Extraction Node

  #13860 opened Feb 17, 2025

• The format of dify's response to the DeepSeek-R1-Distill-Llama-70B model is incorrect. Multiple thinking... collapse tags are displayed.

  #13858 opened Feb 17, 2025

• The mathematical formula of the answer cannot be displayed when using azure deepseek r1 api

  #13857 opened Feb 17, 2025

• build: bump poetry to 2.1

  #13850 opened Feb 17, 2025

• Abnormalities in multi-round LLM conversations

  #13846 opened Feb 17, 2025

• message files are displayed in random order,when browsing history conversation in the logs and annotations page.

  #13841 opened Feb 17, 2025

• enhance chatflow to support options

  #13838 opened Feb 17, 2025

• The message_tokens data in the messages table is abnormal.

  #13837 opened Feb 17, 2025

• Podcast Generator Tool: Missing TTS Voices

  #13822 opened Feb 17, 2025

• The display loading of dropdown options defined as a Constant is problematic

  #13819 opened Feb 17, 2025

• Code execution cannot receive input from LLM in latest version of DIfy.

  #13804 opened Feb 17, 2025

• Add Gitee v5 API as built-in tool

  #13802 opened Feb 17, 2025

• I use DeepSeep R1, Answer will suddenly stop when I ask question

  #13801 opened Feb 17, 2025

• when upgrading from a lower version to a higher version, all existing account data, AI assistant app data, and custom tool data were lost.

  #13791 opened Feb 17, 2025

• Context length for 131072 shoule be 128k maybe?

  #13789 opened Feb 16, 2025

• Input variable for Chat Agent is not working

  #13787 opened Feb 16, 2025

• Support for NIXTLA TIMEGPT

  #13781 opened Feb 16, 2025

• Triggering rate limiting when invoking the AliCloud Rerank model.

  #13780 opened Feb 16, 2025

• Migrate the database and re-vectorization with Error

  #13779 opened Feb 16, 2025

• Ollama call failed.. InternalServerError

  #13777 opened Feb 16, 2025

• Agent unable to read response from custom workflow tools (Anthropic through OpenRouter Amazon Bedrock)

  #13772 opened Feb 15, 2025

• The API key created in a newly established workspace within Aliyun Bailian cannot be configured in Qwen on Dify

  #13769 opened Feb 15, 2025

• In "Create Knowledge Page," during Step One, the section overflows its parent container and does not shrink.

  #13760 opened Feb 15, 2025

• Text to speech not working

  #13759 opened Feb 15, 2025

• Please add support for silicon flow pro models of deepseek-r1 which called Pro/deepseek-ai/deepseek-r1

  #13755 opened Feb 14, 2025

• Fix API Key verification endpoint for SlideSpeak API

  #13754 opened Feb 14, 2025

• Suddenly refresh the page during the chat and ask again for an error

  #13750 opened Feb 14, 2025

• Enhancement of LLM Node Memory Function

  #13738 opened Feb 14, 2025

• add siliconflow apikey error

  #13734 opened Feb 14, 2025

• Plugin Jina.ai cannot be configed

  #13732 opened Feb 14, 2025

• AgentLogEvent is missing `parallel_id` field so it can't run Agent node in parallel.

  #13725 opened Feb 14, 2025

• Suggestions on Model Supplier Interface

  #13724 opened Feb 14, 2025

• Allow Specifying ARN or Custom Inference ID When Selecting "Setup AWS" in Model Provider Settings

  #13710 opened Feb 14, 2025

• Issues with uploading PDF files to the knowledge base

  #13709 opened Feb 14, 2025

• Chat with DeepSeek-R1 API through Azure AI Studio get unexpected end alert

  #13698 opened Feb 13, 2025

• Playwright

  #13697 opened Feb 13, 2025

• details of deepseek-r1's reasoning_content should not return to model for infer

  #13692 opened Feb 13, 2025

• Add new model options for Perplexity Tool (builtin)

  #13688 opened Feb 13, 2025

• dify-web:1.0.0-beta.1 light screenshots not found

  #13687 opened Feb 13, 2025

• Volcengine ark's assistant/bot no response in dify (openapi compatible)

  #13682 opened Feb 13, 2025

• The chat window uploads attachments and sends information to the LLM

  #13680 opened Feb 13, 2025

• Uploaded file name does not match the name in the knowledge base

  #13672 opened Feb 13, 2025

• The language of the 'Next Problem Suggestion' function is incorrect and randomly appears

  #13669 opened Feb 13, 2025

• Unable to delete input variables under specific node combinations in workflow

  #13666 opened Feb 13, 2025

• Question Classfier contains useless content

  #13660 opened Feb 13, 2025

• When calling an LLM in the DIFY workflow, the retry functionality fails when the LLM returns a status of 429.

  #13657 opened Feb 13, 2025

• Error reported when adding text or CSV files to the knowledge base in version 1.0.0

  #13651 opened Feb 13, 2025

• Compatible with Deepseek R1 model access on Xinference

  #13650 opened Feb 13, 2025

• MessageLogModal closed unexpectedly

  #13646 opened Feb 13, 2025

• Variable click jumps to source node

  #13638 opened Feb 13, 2025

• When using Docker Compose to start, the sandbox configuration file isn't loaded from the volume by default.

  #13636 opened Feb 13, 2025

• When I recall knowledge from an external knowledge base API, there is an error, but I can manually adjust the interface through Postman

  #13632 opened Feb 13, 2025

• Streaming Response Broken: Integration of Question Classifier with Conditional Logic Before LLM Node

  #13626 opened Feb 12, 2025

• SMTP/IMAP Email Integration for Email Assistants

  #13625 opened Feb 12, 2025

• Kafka Trigger Integration

  #13624 opened Feb 12, 2025

• When using redis cluster for caching, add global_prefix to all keys cached by redis

  #13604 opened Feb 12, 2025

• support knowledge base for PDF scanned copies

  #13601 opened Feb 12, 2025

• Milvus check_hybrid_search_support Error

  #13595 opened Feb 12, 2025

• Knowledge: retrieval testing cleanup

  #13594 opened Feb 12, 2025

• Knowledge: Retrieval count is always 0

  #13592 opened Feb 12, 2025

• When I open the chat page on my iPhone 13pro max and iPhone 14, the screen is white, but it can be opened on my iPhone 15. Can you help me solve the problem that iPhone13 and iPhone14 cannot be opened?

  #13591 opened Feb 12, 2025

• Why is there no Thinking content in the output of Openrouter's DeepSeek-r1 model?

  #13590 opened Feb 12, 2025

• Suggestions for Enhancements to Support Long Models and System Prompts

  #13586 opened Feb 12, 2025

• Unable to add in batch in the knowledge base

  #13577 opened Feb 12, 2025

• File download failure due to unsupported encoding in Content-Disposition

  #13573 opened Feb 12, 2025

• Duplicate Tool Response Output in Agent Execution

  #13566 opened Feb 12, 2025

• to enhance input efficiency

  #13563 opened Feb 12, 2025

• When using external knowledge bases, it is desired to add additional context parameters

  #13561 opened Feb 12, 2025

• Embedding websites can customize End User

  #13555 opened Feb 12, 2025

• Failed to install canvas as frontend package on Mac

  #13554 opened Feb 12, 2025

• Add reranker - Azure AI Studio

  #13550 opened Feb 12, 2025

• Parent Child (full doc) + Reranking does not seem right

  #13547 opened Feb 11, 2025

• Show the user vote (thumbs-up/down) in logs

  #13539 opened Feb 11, 2025

• Unify Function Call and Tool Call in Agent Mode

  #13533 opened Feb 11, 2025

• When using full-text search or hybrid search, an error occurs during query: [{'locations': [{'column': 92, 'line': 1}], 'message': 'Syntax Error GraphQL request (1:92) Invalid character escape sequence: \\a.\n\n1:'.

  #13529 opened Feb 11, 2025

• thinking block elapsed time error

  #13521 opened Feb 11, 2025

• xinference using local deepseek r1 small models cannot display think tag well

  #13520 opened Feb 11, 2025

• Knowledge base with unstructured integration affects

  #13519 opened Feb 11, 2025

• The DeepSeek R1 model deployed on Azure cannot be accessed.

  #13517 opened Feb 11, 2025

81 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• resolved: add basic auth for chatbot , implemented basic authentication after deploying with Docker Compose

  #10215 commented on Feb 15, 2025 • 2 new comments

• Introduce Configurable Upper-Limits/default values for max retries and retry Interval in Error Handling(#12787)

  #12788 commented on Feb 13, 2025 • 1 new comment

• fix: fail-branch stream output error

  #13401 commented on Feb 11, 2025 • 0 new comments

• Custom Tools With multipart/forms Content

  #12800 commented on Feb 17, 2025 • 0 new comments

• Mermaid rendering error due to line break(both "<br>" and "\n") in node content.

  #13481 commented on Feb 17, 2025 • 0 new comments

• The logs and annotations fail to display the dialogue content

  #12528 commented on Feb 17, 2025 • 0 new comments

• API Timeout Issue in a Self-Hosted Web Workflow

  #12834 commented on Feb 17, 2025 • 0 new comments

• Can the admin see which specific messages the users have provided feedback on?

  #11576 commented on Feb 17, 2025 • 0 new comments

• when i upload file,Text clean the file ,the file size is zero

  #12813 commented on Feb 17, 2025 • 0 new comments

• The variable output by the code snippet in the workflow cannot be found in the LLM node

  #12824 commented on Feb 17, 2025 • 0 new comments

• Among the built-in tools, Feishu provides the send_bot_message tool, but a 404 error is reported when accessing the interface.

  #12791 commented on Feb 17, 2025 • 0 new comments

• Request: API - Body/Content type | Text (text/plain)

  #12831 commented on Feb 17, 2025 • 0 new comments

• Make.com integration

  #12832 commented on Feb 17, 2025 • 0 new comments

• Add default values to user-defined input field

  #11290 commented on Feb 16, 2025 • 0 new comments

• add Crawl4AI to tools and for sync data from website in Knowledge

  #8905 commented on Feb 16, 2025 • 0 new comments

• Issue with chunk splitting using parent-child mode in a Markdown file

  #13132 commented on Feb 15, 2025 • 0 new comments

• Request to Add API Base Configuration for Google Gemini Model Provider

  #13047 commented on Feb 15, 2025 • 0 new comments

• Endpoint for insert data in memory

  #12743 commented on Feb 15, 2025 • 0 new comments

• When adding a knowledge base to the Workflow-Knowledge Retrieval node, it is recommended to add a search function to quickly retrieve the required knowledge base and improve efficiency. See screenshot for details

  #11372 commented on Feb 15, 2025 • 0 new comments

• Got validation failed error for custom file extension in Chatflow

  #13148 commented on Feb 14, 2025 • 0 new comments

• Suggested questions always return empty

  #11818 commented on Feb 14, 2025 • 0 new comments

• Support for Vendor-Specific Built-in Function Calls in Dify Models

  #12710 commented on Feb 14, 2025 • 0 new comments

• feat: Support deployment in custom subpaths by `NEXT_PUBLIC_BASE_PATH`

  #13394 commented on Feb 14, 2025 • 0 new comments

• chore: Bump ruff to 0.9.6

  #13356 commented on Feb 17, 2025 • 0 new comments

• Update .env.example to fix MILVUS_URI default value

  #13140 commented on Feb 12, 2025 • 0 new comments

• fix parent-child retrieval return nothing in agent mode without rerank

  #13084 commented on Feb 13, 2025 • 0 new comments

• fix: run out of memory due to indexing/embedding documents

  #12882 commented on Feb 11, 2025 • 0 new comments

• Feature/sambanovacloud integration dify

  #11918 commented on Feb 13, 2025 • 0 new comments

• feat: add support for Azure speech

  #11375 commented on Feb 11, 2025 • 0 new comments

• Statistic date range 'Today' send invalid range

  #13105 commented on Feb 18, 2025 • 0 new comments

• Application erroriaclientside exception has occurred (see the browser console for moreinformation), Error opening the application using UOS arm browser

  #13402 commented on Feb 17, 2025 • 0 new comments

• Clarification of software architecture, namely pluginDaemon

  #12869 commented on Feb 17, 2025 • 0 new comments

• LLM node receives multiple base64-encoded images

  #12398 commented on Feb 17, 2025 • 0 new comments

• yahoo tools is not avaliable

  #13403 commented on Feb 17, 2025 • 0 new comments

• Attempting to use search tools results in bad error request 400

  #13254 commented on Feb 17, 2025 • 0 new comments

• /chat-messages (service api) `files` parameter transfer_method to add `tool file` type

  #12777 commented on Feb 17, 2025 • 0 new comments

• Implement OpenAI's Realtime API

  #8986 commented on Feb 17, 2025 • 0 new comments

• Calling the interface for choosing the tongyi model results in an error when selecting blocking mode.

  #13512 commented on Feb 17, 2025 • 0 new comments

• How to solve one sentence is split into two chunks when parsing PDF files?

  #13378 commented on Feb 17, 2025 • 0 new comments

• Disconnection of network for workflow execution causes status stock as 'running' due to main thread kill while actual execution completes at backend thread

  #12798 commented on Feb 17, 2025 • 0 new comments

• Confluence and Jira Integration for Streamlined Workflows

  #9805 commented on Feb 17, 2025 • 0 new comments

• Can the input variables include a feature to display conditionally based on certain criteria?

  #12554 commented on Feb 13, 2025 • 0 new comments

• Can a feature be added to allow redoing word embedding for specific chunks that failed during the knowledge base embedding process?

  #12524 commented on Feb 13, 2025 • 0 new comments

• Perpetual queue for uploading a document to the knowledge base

  #11911 commented on Feb 13, 2025 • 0 new comments

• HTTP Request Node Not Retrieve File Extension from Content-Disposition Header

  #12608 commented on Feb 12, 2025 • 0 new comments

• 1.0.0-beta.1-Internal ServerError

  #12617 commented on Feb 12, 2025 • 0 new comments

• Add support for import some Q&A text not need LLM to generate results

  #12842 commented on Feb 12, 2025 • 0 new comments

• Support for Publishing Multiple Applications on a Single Platform for Users

  #13513 commented on Feb 12, 2025 • 0 new comments

• The AI Web App shows a blank page when opened on iPhone X.

  #12019 commented on Feb 12, 2025 • 0 new comments

• Dynamically choose model for LLM node

  #13298 commented on Feb 12, 2025 • 0 new comments

• Firecrawl and Jina Reader APIs don't work

  #13081 commented on Feb 12, 2025 • 0 new comments

• Add WHILE loop or TRY block to workflow

  #12289 commented on Feb 12, 2025 • 0 new comments

• Add Japanese documentation(`template.ja.mdx`) for Knowledge API

  #12359 commented on Feb 12, 2025 • 0 new comments

• The "dify-client" npm package is 7 months old

  #12183 commented on Feb 12, 2025 • 0 new comments

• Unable to Pass Parameters When Embedding Dify Chat Interface via iframe

  #12384 commented on Feb 12, 2025 • 0 new comments

• Enhancing Parent-Child Retrieval by Adding Keywords to Child Blocks

  #12630 commented on Feb 12, 2025 • 0 new comments

• Add Arize OSS Phoenix integration

  #12140 commented on Feb 11, 2025 • 0 new comments

• Add "is null" operator in "IF/ELSE" block

  #12646 commented on Feb 11, 2025 • 0 new comments

• Add SambaNova as a model provider

  #13470 commented on Feb 11, 2025 • 0 new comments

• Knowledge description character limit discrepancy

  #13503 commented on Feb 11, 2025 • 0 new comments

• Application error: a client-side exception has occurred (see the browser console for more information).

  #12198 commented on Feb 11, 2025 • 0 new comments

• The knowledge base needs to be able to add custom fields and use them for filtering in workflow knowledge retrieval, which would be a particularly useful feature

  #9630 commented on Feb 14, 2025 • 0 new comments

• Using a custom image for the copied app icon does not work as expected

  #12732 commented on Feb 14, 2025 • 0 new comments

• web page lack full text search and hybrid search

  #12730 commented on Feb 14, 2025 • 0 new comments

• Llama3.1 serial models in Azure AI Studio don't work

  #12742 commented on Feb 14, 2025 • 0 new comments

• Deepseek r1 deployed using ollama cannot use the network search tool in the dify studio

  #13468 commented on Feb 14, 2025 • 0 new comments

• The Japanese PDF gets garbled when registering it as knowledge.

  #12457 commented on Feb 14, 2025 • 0 new comments

• Uploaded files accumulating in '\volume\app\storage\uploaded_files'

  #12433 commented on Feb 14, 2025 • 0 new comments

• The LLM node following multiple consecutive IF-ELSE nodes cannot perform streaming output normally.

  #13264 commented on Feb 14, 2025 • 0 new comments

• Support grounding for Vertex Gemini

  #12990 commented on Feb 14, 2025 • 0 new comments

• DuckDuckGo Search returns "subsection not found"

  #13458 commented on Feb 14, 2025 • 0 new comments

• The file upload functionality does not support non-image file formats.

  #11652 commented on Feb 14, 2025 • 0 new comments

• Query Transformations for Improved Retrieval in RAG

  #12660 commented on Feb 13, 2025 • 0 new comments

• Deploying the latest version and integrating the Qwen model results in an error.

  #12681 commented on Feb 13, 2025 • 0 new comments

• When copying an iteratation node in a workflow, the edges between the nodes inside the copied iteration node are lost.

  #12689 commented on Feb 13, 2025 • 0 new comments

• Naming of Variable Aggregator

  #12628 commented on Feb 13, 2025 • 0 new comments

• Output of Variable Aggregator (Grouping) Not Properly Resolved in Custom External Tool within Workflow

  #12686 commented on Feb 13, 2025 • 0 new comments

• When opening "Model Providers" the page displays an error.

  #12670 commented on Feb 13, 2025 • 0 new comments

• LLM VISION capability CANNOT apply the HTTPS Request data to recognize

  #12685 commented on Feb 13, 2025 • 0 new comments

• Feature Request: Configurable Concurrency for Workflow Batch Runs

  #12906 commented on Feb 13, 2025 • 0 new comments

• In the recall test of the new knowledge base, the number of recall results is incorrect, and the scores of different rerank models are completely identical.

  #13336 commented on Feb 13, 2025 • 0 new comments