OptionalBoundaryTranscript: a silence gap at least this long starts a new chapter.
OptionalEmitTranscript: also emit one child segment per speaker turn within each chapter.
OptionalMaxTranscript: maximum wall-clock length of one chapter, in milliseconds.
OptionalMaxHard ceiling on tokens per segment. Segments larger than this are split.
OptionalMinMerge adjacent text segments estimating below this many tokens.
OptionalMinSemanticText: skip the LLM boundary pass for documents below this token count.
OptionalNoAdaptiveBoundary: if the whole text is within this percent of target, don't split at all.
OptionalOverlapOverlap tokens applied when an oversized segment must be split.
OptionalOvershootAdaptiveBoundary: percent above target to keep looking for a sentence/word break.
OptionalTargetAdaptiveBoundary: desired segment size — size this to your queries, not to the model.
OptionalUndershootAdaptiveBoundary: percent below target at which a paragraph break is accepted.
OptionalWindowFixedWindow: window length in milliseconds for audio/video with no transcript.
Options for the segmentation strategy. All optional; each segmenter ignores options that don't apply to it.