Google Gemini 2.0 flash has wider range of temperature.

[anilgulecha]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain.js documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain.js rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

const baseLLM = new ChatGoogleGenerativeAI({
model: "gemini-2.0-flash",
temperature: 1,
maxRetries: 1,
maxConcurrency: 1
});

Error Message and Stack Trace (if applicable)

Error: temperature must be in the range of [0.0,1.0]
at new ChatGoogleGenerativeAI

Description

https://github.com/langchain-ai/langchainjs/blob/main/libs/langchain-google-genai/src/chat_models.ts#L647

The code assumes the temperature has to be b/w 0 and 1. But the latest model supports upto 2 (gemini-2.0-flash)

System Info

I'm on

    "@langchain/core": "^0.3.39",
    "@langchain/google-genai": "^0.1.8",
    "@langchain/langgraph": "^0.2.46"

[dosubot]

Hello, @anilgulecha! I'm here to assist you with bugs, questions, and becoming a contributor. Let's squash those bugs together! 🤖

To resolve the issue with the gemini-2.0-flash model's temperature parameter, you need to update the validation logic in the LangChain.js code. The current implementation enforces a range of [0.0, 1.0], but the gemini-2.0-flash model supports values up to 2. You can modify the code as follows:

In the constructor of the GoogleGenerativeAIChatInput class, update the condition to:

this.temperature = fields?.temperature ?? this.temperature;
if (this.temperature && (this.temperature < 0 || this.temperature > 2)) {
  throw new Error("`temperature` must be in the range of [0.0,2.0]");
}

This change will allow the temperature parameter to accept values up to 2, accommodating the gemini-2.0-flash model's requirements ^[1][2].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[afirstenberg]

You may also wish to switch to the @langchain/google-webauth package which has this change already applied.

[anilgulecha]

Hi How can I pull this locally? currently I see only 0.1.8 available in npm
https://www.npmjs.com/package/@langchain/google-genai

When will newer version be tagged?

Regards

[jacoblee93]

Ah I think this was only fixed for vertex, need to do another release for Gen AI!

[afirstenberg]

Ah I think this was only fixed for vertex, need to do another release for Gen AI!

This had been fixed in Vertex/common for a while.
These changes were for genai.
(They just came at the same time I did a bunch of other work for Vertex)

I think all you need to do is cut a new release for google-genai.

[jacoblee93]

Oh my mistake! Will do later today

[jacoblee93]

Sorry about the delay, just cut!