Token Limits
ArchiveVery Easy
A response is cut off mid-sentence. The context window isn't full, but the output stops abruptly. (Note: Google calls this max_output_tokens.)
What parameter caps the length of the generated response?
Flag format: CONGRESS{[parameter_name]}
Example: CONGRESS{temperature}
Show hint
This parameter limits how many tokens the model will generate.
Archive — no submissions accepted
This challenge is preserved for reference. Play live challenges at /challenges.