gemini-2.0-flash-thinking-exp-1219 does not work

[andresgit]

What happened?

Trying gemini-2.0-flash-thinking-exp-1219 gives:
"Something went wrong. Here's the specific error message we encountered: An error occurred while processing your request. Please contact the Admin."

Steps to Reproduce

1. try gemini-2.0-flash-thinking-exp-1219

What browsers are you seeing the problem on?

No response

Relevant log output

No response

Screenshots

Code of Conduct

• I agree to follow this project's Code of Conduct

[danny-avila]

It works, but for some reason Google has made one of the safety settings a requirement.

Besides adding the model to the list, you need to add the following to your .env file:

GOOGLE_SAFETY_CIVIC_INTEGRITY=BLOCK_NONE

Or whichever safety setting you prefer:
https://ai.google.dev/gemini-api/docs/safety-settings

[danny-avila]

I'll make BLOCK_NONE the default value for this if not provided in the .env

[wsybok]

Gemini's new "thinking-exp" model appears to generate a detailed thought process, while LibreChat functions as expected, producing standard output and counting as output tokens. I believe this difference is due to the use of the langchain package. Perhaps we should consider migrating to the Google GenAI SDK?

[danny-avila]

This is expected.

it is actually using @google/generative-ai in this case
const { GoogleGenerativeAI: GenAI } = require('@google/generative-ai');

[danny-avila]

curl is the same

[
    {
        "candidates": [
            {
                "content": {
                    "parts": [
                        {
                            "text": "Here's a breakdown of my thought process in generating the Python list sorting code"
                        }
                    ],
                    "role": "model"
                },
                "index": 0,
                "safetyRatings": [
                    {
                        "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT",
                        "probability": "NEGLIGIBLE"
                    },
                    {
                        "category": "HARM_CATEGORY_HATE_SPEECH",
                        "probability": "NEGLIGIBLE"
                    },
                    {
                        "category": "HARM_CATEGORY_HARASSMENT",
                        "probability": "NEGLIGIBLE"
                    },
                    {
                        "category": "HARM_CATEGORY_DANGEROUS_CONTENT",
                        "probability": "NEGLIGIBLE"
                    }
                ]
            }
        ],
        "usageMetadata": {
            "promptTokenCount": 9,
            "candidatesTokenCount": 16,
            "totalTokenCount": 25
        },
        "modelVersion": "gemini-2.0-flash-thinking-exp-1219"
    },
    {
        "candidates": [
            {
                "content": {
                    "parts": [
                        {
                            "text": " and explanation:\n\n1. **Understanding the Core Request:** The user wants to"
                        }
                    ],
                    "role": "model"
                },