Setting the maxOutputTokens with ChatVertexAI to maximum available

[ryan65]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

this.model = new ChatVertexAI(config);

Description

Hi I am using typescript langchain , and creating an instance of ChatVertextAI.
I cant figure out the maxOutputTokens parameter.
I would like to avoid setting it and having the model always use the maximum tokens allowed for example with Gemini 2.0 flash its 8192.
since I have different models I dont want to store the max tokens of each llm. and would prefer to always use the maximum available.
Is there an easy way ?. I see that if I dont define maxOutputTokens the response is sometimes cut off . I would have expected the behaviour to be that if maxOutputTokens is not defined use the maximum.
Question: does langchain library in anyway set a default (couldnt find any mention but still asking)
if its not langchain so is this the default of the model itself when maxOutputTokens is not defined.?
Thanks.

System Info

using langchain 0.3.12 and

[ryan65]

Just to clarify.
According the documentation I found , without setting the maxOutputTokens , it should be by default 8192 but I see this is not the case. when not setting it , a specific response is cut in the middle and when configuring it to 8192 I get the full response. so it seems that this does have an effect.