[Support]: Facing context_length_exceeded Error with Claude-4.5-Haiku Due to Message Token Limit

[mansiibm]

Hi, I was using the anthropic model- claude-4.5-haiku which has the token limit of -

some 200k tokens so I got the following warning as -

{"attemptNumber":1,"code":"context_length_exceeded","error":{"code":"context_length_exceeded","message":"This model's maximum context length is 128000 tokens. However, your messages resulted in 167630 tokens. Please reduce the length of the messages.","param":"messages","type":"invalid_request_error"},"headers":{}

I have not set any limits for the tokens for my anthropic model, still I face this issue.

I am using the LC version - v0.8.0. How can I avoid having this issue? Occurs for other models also.

[danny-avila]

You need to adjust "max output tokens" since by default it is 64,000, that should help.

[mansiibm]

Hi Danny, thankyou for replying , but if I set the maxOutputTokens- it will be part of the preset right ? I do not wish to create a model spec for this model only. Can we please have some config/setting without preset just against the endpoint or even something like this -

anthropic: 
    models:
          haiku-4-5-latest: 
              maxOutputTokens: 200000

Please correct me if I am wrong, either I can set it in the form of a model spec or in the UI as- preset right, there is not in between, sorry is it is already answered somewhere. Thank you and good day!!