Multimodal LLM analyze image errors.

[DEMOUSL]

Self Checks

• I have read the Contributing Guide and Language Policy.
• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report, otherwise it will be closed.
• 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）
• Please do not modify this template :) and fill in all the required fields.

Dify version

1.11.4

Cloud or Self Hosted

Cloud

Steps to reproduce

When I tried to analyze images or videos, an error occurred. I tried different ways of entering the parameters, and it only worked with plain text. Errors occurred when using multimodal input.

✔️ Expected Behavior

The analysis results, including images or videos, can be output normally.

❌ Actual Behavior

No response

[dosubot]

Hi @DEMOUSL! I'm Dosu and I’m helping the dify team.

This is a known bug in Dify v1.11.4 (cloud) affecting multimodal LLM workflows, especially with Gemini 3 Pro Preview. The error happens because the system can't generate or retrieve the required file upload URL for images or videos during plugin invocation, which breaks multimodal analysis—plain text works since it doesn't need file upload handling. The root cause is a recent change that deleted the /datasets/{dataset_id}/documents/{document_id}/upload-file API endpoint, which previously provided upload file URLs for plugins and workflows. Without this, the backend can't validate or pass files to the model, resulting in no output or a JSON error about missing upload URLs source.

There’s no official fix yet. Workarounds include switching to models that don’t require strict file upload handling (like Qwen-Max or DeepSeek), converting videos to images or audio for analysis, or using plugin versions that pin older, compatible SDKs if available source. For video, converting to images or audio is the most reliable option right now source.

You can track progress or updates in the open issue here. If your scenario differs or you find a new edge case, sharing details in the issue may help the maintainers.

To reply, just mention @dosu.

---

Share context across your team and agents. Try Dosu.

[QuantumGhost]

@DEMOUSL Could you please provide a minimal reproducible example to help us troubleshooting this problem?

[DEMOUSL]

image analysis.yml
This is my most simplified method for generating multimodal output.

[fatelei]

Self Checks

• I have read the Contributing Guide and Language Policy.[x] This is only for bug report, if you would like to ask a question, please head to Discussions.[x] I have searched for existing issues search for existing issues, including closed ones.[x] I confirm that I am using English to submit this report, otherwise it will be closed.[x] 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）[ ] Please do not modify this template :) and fill in all the required fields.

Dify version

1.11.4

Cloud or Self Hosted

Cloud

Steps to reproduce

When I tried to analyze images or videos, an error occurred. I tried different ways of entering the parameters, and it only worked with plain text. Errors occurred when using multimodal input.

✔️ Expected Behavior

The analysis results, including images or videos, can be output normally.

❌ Actual Behavior

No response

need upgrade the plugin, after plugin updated, there is parameter use_inline_file, enable this parameter, it will not upload file to google

[fatelei]

this error is upload file to google, it failed

[QuantumGhost]

Gemini plugin 0.7.4 works for me.

@DEMOUSL Could you please upgrade you gemini plugin to 0.7.4 and check if the problem still exists?

[DEMOUSL]

@fatelei @QuantumGhost I have updated the plugin to version 0.7.4, but the error persists, and I couldn't find the new use_inline_file parameter in the settings. Is there anything else I need to modify?

[fatelei]

@fatelei @QuantumGhost I have updated the plugin to version 0.7.4, but the error persists, and I couldn't find the new use_inline_file parameter in the settings. Is there anything else I need to modify?

currently only model gemini-3-flash-preview support this parameter

[DEMOUSL]

@fatelei I tried gemini-3-flash-preview and it worked, but the results seemed slightly worse than with the pro version. Could you please add the same parameters to gemini-3-pro-preview later?

[fatelei]

@fatelei I tried gemini-3-flash-preview and it worked, but the results seemed slightly worse than with the pro version. Could you please add the same parameters to gemini-3-pro-preview later?

already submit a pr langgenius/dify-official-plugins#2439