[BUG] Gemini Ai integration doesn't work

[Delivator]

• I confirm that I have reviewed the FAQ, issues, and discussions, and found no existing solution.

Describe the bug

Using the Instant Playlist feature results in an instant error. In the web ui it says:

Built the playlist0.1s
The AI couldn't build a plan for this request - try rephrasing it.

To Reproduce

1. Set up the Ai provider using the setup wizard (AI_MODEL_PROVIDER = GEMINI, GEMINI_API_KEY double checked and updated several times, GEMINI_MODEL_NAME=gemini-3.5-flash (I have also tried gemini-2.5-flash and gemini-2.5-pro)

Expected behavior

It works.

Environment (please complete the following information):

• OS: Linux/TrueNAS 25.04.2.6
• Deployment: Docker/TrueNAS Custom YAML Config
• AudioMuse-AI Version: latest (v2.6.0)
• Navidrome/Jellyfin Version: Jellyfin v10.11.11
• Browser: Firefox

Additional context

Log from the audiomuse-ai-flask container:

[INFO]-[10-07-2026 10-01-02]-chat_playlist_api -> provider=GEMINI url=(gemini-api) model=gemini-3.5-flash (default_provider=GEMINI, client_override=True)
[ERROR]-[10-07-2026 10-01-02]-Error calling Gemini with tools
Traceback (most recent call last):
  File "/app/tasks/ai/providers/gemini.py", line 103, in call_with_tools
    tools_list = [genai.types.Tool(function_declarations=function_declarations)]
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/dist-packages/pydantic/main.py", line 263, in __init__
    validated_self = self.__pydantic_validator__.validate_python(data, self_instance=self)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
pydantic_core._pydantic_core.ValidationError: 5 validation errors for Tool
function_declarations.3.parameters.properties.genres.uniqueItems
  Extra inputs are not permitted [type=extra_forbidden, input_value=True, input_type=bool]
    For further information visit https://errors.pydantic.dev/2.13/v/extra_forbidden
function_declarations.3.parameters.properties.voices.uniqueItems
  Extra inputs are not permitted [type=extra_forbidden, input_value=True, input_type=bool]
    For further information visit https://errors.pydantic.dev/2.13/v/extra_forbidden
function_declarations.3.parameters.properties.moods.uniqueItems
  Extra inputs are not permitted [type=extra_forbidden, input_value=True, input_type=bool]
    For further information visit https://errors.pydantic.dev/2.13/v/extra_forbidden
function_declarations.3.parameters.properties.exclude_artists.uniqueItems
  Extra inputs are not permitted [type=extra_forbidden, input_value=True, input_type=bool]
    For further information visit https://errors.pydantic.dev/2.13/v/extra_forbidden
function_declarations.3.parameters.properties.exclude_genres.uniqueItems
  Extra inputs are not permitted [type=extra_forbidden, input_value=True, input_type=bool]
    For further information visit https://errors.pydantic.dev/2.13/v/extra_forbidden
172.16.9.1 - - [10/Jul/2026:10:01:02 +0000] "POST /chat/api/chatPlaylistStream HTTP/1.1" 200 1781 "https://audiomuse-ai.harz.lol/chat/" "Mozilla/5.0 (X11; Linux x86_64; rv:152.0) Gecko/20100101 Firefox/152.0"

[NeptuneHub]

Hi and thanks for the feedback. In this PR I added the fix:

• AI tools integration fix #747

Can you wait around 30-60 minutes to have the new PR imaged builded (you will find the name attached to the PR when complete), give it a try and confirm if this work?

Thanks.

[Delivator]

I've tried the tagged image and the issue seems to be fixed.

[NeptuneHub]

Great, thanks for discovering it and for the testing. Going to close the issue.

[github-actions]

We've closed the issue now, thanks for having starred AudioMuse-AI and supporting self-hosted music here on GitHub! It really helps the project grow!