v0.15.2

[laipz8200]

✨ What’s New in v0.15.2? ✨

The v0.15.2 release is all about squashing bugs, unlocking new capabilities, and generally making things run smoother. Here's a peek at what's changed:

🚀 New Features

• Minimax-Text-01 Support: Added support for the new Minimax-Text-01 model by @AAlexDing in feat:Support Minimax-Text-01 #12763.
• Gemini-2.0-Flash-Thinking-Exp-01-21: New experimental model added by @jiandanfeng in feat: add gemini-2.0-flash-thinking-exp-01-21 #12924.
• Doubao 1.5 Model Series: Support for Doubao 1.5 models in our tooling setup by @sinomoe in feat(ark): support doubao 1.5 series of models #12935.
• Profile Avatar Customization: You can now change your profile avatar and keep it fresh, thanks to @miya in feat: Add ability to change profile avatar #12642.
• Environment Configuration for Chunk Length: Configure document chunk length through environment variables by @iamjoel in feat: support config chunk length by env #12925.

🛠️ Enhancements

• Credential Extraction Logic: Enhanced credential extraction logic for better configurability by @fujita-h in feat: enhance credential extraction logic based on configurate method #12853.
• Document Chunk Setting Updates: Now possible to update chunk settings for existing documents by @kurokobo in feat: allow updating chunk settings for the existing documents #12833.
• Full-Text Search in Milvus: You can now configure full-text search capabilities within Milvus by @JohnJyong in add milvus full text search setting #12930.

🐞 Bug Fixes

• Rerank Model Bug: Fixed the rerank model and switch issues across various tools by @WTW0313 in fix: Fix rerank model switching issue #12721 and @luckylhb90 in bug: fixed bedrock rerank bug #12774.
• Clipboard WriteText on HTTP: Resolved issues with clipboard text writing under the HTTP protocol by @le0zh in fix: resolve clipboard.writeText failure under HTTP protocol #12936.
• OpenAI Bad Request Error: Cleaned up issues causing bad request errors on OpenAI o1 endpoint by @k-zaku in fix: OpenAI o1 Bad Request Error #12839.
• ReAct Inference Error: Fixed error logging issue when ReAct inference mode is enabled by @jiandanfeng in fix: Agent is configured for ReAct inference mode, an error is reported when viewing the agent log #12920.
• Agent Log Formatting: Tweaked DeepSeek's API response handling to fix format issues by @jiandanfeng in fix: DeepSeek API Error with response format active (text and json_object)  #12747.

🧑‍💻 Developer Note

Note

Poetry Upgrade Alert: Heads up, devs! We've upgraded to Poetry 2.x from 1.x. This upgrade is aimed at making your package management smoother and more efficient, so make sure to update your local setup and check out what's new in Poetry 2.x by @bowenliang123 in #12369.

So, what's next? Get this version in your dev arsenal and explore these cool upgrades in Dify! As always, dive in, have fun, and let us know your thoughts.

---

Upgrade Guide

Docker compose deployments

Warning

The files in the docker-legacy directory will soon stop being maintained and will be removed from the repository. If you are still using them, please switch to the new version as soon as possible.

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

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

2. Get the latest code from the main branch

   git checkout main
   git pull origin main

3. Stop the service，Command, please execute in the docker directory

   docker compose down

4. Back up data

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

5. Upgrade services

   docker compose up -d

Source Code deployments

1. Stop API server, Worker and Web frontend Server.

2. Get the latest code from the release branch:

   git checkout 0.15.2

3. Update Python dependencies:

   cd api
   poetry install

4. Then, let's run the migration script:

   poetry run flask db upgrade

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

---

What's Changed

• Feat: copyright modification by @JzoNgKVO in Feat: copyright modification #12707
• chore: translate i18n files by @github-actions in chore: translate i18n files #12708
• fix: Fix rerank model switching issue by @WTW0313 in fix: Fix rerank model switching issue #12721
• feat: enhance credential extraction logic based on configurate method by @fujita-h in feat: enhance credential extraction logic based on configurate method #12853
• fix: DeepSeek API Error with response format active (text and json_object) by @jiandanfeng in fix: DeepSeek API Error with response format active (text and json_object)  #12747
• chore: fix chinese translation for 'recall' by @xhdd123321 in chore: fix chinese translation for 'recall' #12772
• fix(i18n): correct typo in zh-Hant translation by @yenslife in fix(i18n): correct typo in zh-Hant translation #12852
• Fix suggested_question_prompt by @xu-song in Fix suggested_question_prompt #12738
• fix: "parmas" spelling mistake. by @jiandanfeng in fix: "parmas" spelling mistake. #12875
• fix: SparkLite API Auth error (SparkLite API Auth error #12781) by @jiandanfeng in fix: SparkLite API Auth error (#12781) #12790
• feat: allow updating chunk settings for the existing documents by @kurokobo in feat: allow updating chunk settings for the existing documents #12833
• fix: serply credential check query might return empty records by @rayshaw001 in fix: serply credential check query might return empty records #12784
• feat:Support Minimax-Text-01 by @AAlexDing in feat:Support Minimax-Text-01 #12763
• chore(lint): fix quotes for f-string formatting by bumping ruff to 0.9.x by @bowenliang123 in chore(lint): fix quotes for f-string formatting by bumping ruff to 0.9.x #12702
• fix: Issues related to the deletion of conversation_id (Issues related to the deletion of conversation_id #12488) by @jiandanfeng in fix: Issues related to the deletion of conversation_id (#12488) #12665
• chore(fix): Invalid quotes for using Array[String] in HTTP request node as JSON body by @bowenliang123 in chore(fix): Invalid quotes for using Array[String] in HTTP request node as JSON body #12761
• add deepseek-reasoner by @wangiii in add deepseek-reasoner #12898
• fix: external dataset hit test display issue(Test external knowledge base retrieval, the console reports an error in v0.15.0 #12564) by @le0zh in fix: external dataset hit test display issue(#12564) #12612
• Update deepseek model configuration by @wangiii in Update deepseek model configuration #12899
• fix: OpenAI o1 Bad Request Error by @k-zaku in fix: OpenAI o1 Bad Request Error #12839
• fix: variable panel scrollable by @xhdd123321 in fix: variable panel scrollable #12769
• chore: remove webapp ga by @iamjoel in chore: remove webapp ga #12909
• bug: fixed bedrock rerank bug by @luckylhb90 in bug: fixed bedrock rerank bug #12774
• fix update segment keyword with same content by @JohnJyong in fix update segment keyword with same content #12908
• fix:Fix a bug that returns null when the passed path is a file. by @jiangbo721 in fix:Fix a bug that returns null when the passed path is a file. #12775
• feat: Add ability to change profile avatar by @miya in feat: Add ability to change profile avatar #12642
• feat: add gemini-2.0-flash-thinking-exp-01-21 by @jiandanfeng in feat: add gemini-2.0-flash-thinking-exp-01-21 #12924
• feat: support config chunk length by env by @iamjoel in feat: support config chunk length by env #12925
• fix: app token's last_used_at can't be updated when last_used_at is null by @Egfly in fix: app token's last_used_at can't be updated when last_used_at is null #12770
• fix: Agent is configured for ReAct inference mode, an error is reported when viewing the agent log by @jiandanfeng in fix: Agent is configured for ReAct inference mode, an error is reported when viewing the agent log #12920
• add milvus full text search setting by @JohnJyong in add milvus full text search setting #12930
• chore(build): bump poetry from 1.x to 2.x by @bowenliang123 in chore(build): bump poetry from 1.x to 2.x #12369
• fix(batch_create_segment_to_index_task): count max_position in memory. by @laipz8200 in fix(batch_create_segment_to_index_task): count max_position in memory. #12929
• feat: workflow note dark theme by @zxhlyh in feat: workflow note dark theme #12932
• fix: resolve clipboard.writeText failure under HTTP protocol by @le0zh in fix: resolve clipboard.writeText failure under HTTP protocol #12936
• feat(ark): support doubao 1.5 series of models by @sinomoe in feat(ark): support doubao 1.5 series of models #12935
• chore: update version to 0.15.2 in packaging and docker configurations by @laipz8200 in chore: update version to 0.15.2 in packaging and docker configurations #12940

New Contributors

• @jiandanfeng made their first contribution in fix: DeepSeek API Error with response format active (text and json_object)  #12747
• @yenslife made their first contribution in fix(i18n): correct typo in zh-Hant translation #12852
• @rayshaw001 made their first contribution in fix: serply credential check query might return empty records #12784
• @le0zh made their first contribution in fix: external dataset hit test display issue(#12564) #12612
• @k-zaku made their first contribution in fix: OpenAI o1 Bad Request Error #12839
• @Egfly made their first contribution in fix: app token's last_used_at can't be updated when last_used_at is null #12770

Full Changelog: 0.15.1...0.15.2

---

This discussion was created from the release v0.15.2.

[akonnerr]

Why is the version still 15.1 instead of 15.2 when I use git checkout main and git pull origin main to upgrade?

[bowenliang123]

I guess your local main branch is using your own remote repo as the upstream, while dify remote repo should be the correct one.
The version in the code clearly updated as 0.15.2. source code:
https://github.com/langgenius/dify/blob/0.15.2/api/configs/packaging/__init__.py#L12