Response Length Limit - Words/Tokens

Is there a limit on the response length returned?

I’m seeing some examples where the response text is cut off. Presumably, because the response length is too long. Example attached.

If there is a specific response length I will tell my prompt to limit the length.


There are indeed limits on the response length. This is a configurable option inside the no-code AI builder. It’s under the Advanced Options and labeled as “Maximum Output Length”. You set this limit to be whatever you want. Importantly, this does not mean the Pickaxe will always aim for that length (that is best controlled in the prompt), but it will not exceed that length (i.e. that much space will be reserved for the output). The upper limit of this slider is the limit of the underlying model itself.

Here is a screenshot of where to set response length.

1 Like

Yeah I’m having this same issue. Despite specifying multiple times in the prompt that the output should be 2500-words and setting the Max’ Output Length to its maximum 4095 tokens. I am still seeing outputs cut off before completion (after just 800 words or so). No idea why, please can you help?

1 Like