embedding — A list of numbers representing a piece of text's meaning, so that similar meanings end up numerically close and can be searched.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
Why it matters
EVIE-4.5B lets you truncate a single embeddingA list of numbers representing a piece of text's meaning, so that similar meanings end up numerically close and can be searched.Full definition → to 64-2048 dimensions at runtime and compresses ~750 page tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → down to 32 via training-free clustering, shrinking multi-vector document retrieval indexes to under 4GB per million pages while topping ViDoRe V3.
