Getting PROHIBITED_CONTENT error with Gemini 3 pro API

Gemini API
,
1

Getting the output as:

{
“candidates”: [
{
“content”: {}
}
],
“promptFeedback”: {
“blockReason”: “PROHIBITED_CONTENT”,
“blockReasonMessage”: “The model could not generate images because the input image violates Google’s [Generative AI Prohibited Use policy](https://policies.google.com/terms/generative-ai/use-policy). If you think this was an error, [send feedback](https://ai.google.dev/gemini-api/docs/troubleshooting).”
},
“usageMetadata”: {
“promptTokenCount”: 1094,
“totalTokenCount”: 1094,
“promptTokensDetails”: [
{
“modality”: “TEXT”,
“tokenCount”: 1094
}
]
},
“turnToken”: “v1_ChdQb013YWEyYkhLblhqdU1QaGM2MDJBRRIXUG9Nd2FhMmJIS25YanVNUGhjNjAyQUU”
}

This happens during generating text itself, paraphrasing a single line of prompt sometimes fixes it, however, often it comes during the video understanding task.

3

Hi @Parakh_Agarwal , Welcome to forum!!!

You’re encountering a Safety Filter Block because Gemini flagged part of your input often a single video frame as potentially unsafe, causing the whole request to fail. Since the API’s safety filters are strict and can over-trigger, refining or simplifying your prompt can help.For more details, please check the referenced documentation.

Thanks.