OpenRouter's May 2025 Cerebras provider announcement advertised more than 3,000 generated 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 → per second for Llama 4 Scout and more than 1,800 for Llama 3.3 70B Instruct. These figures describe text-generation throughput for particular models. For application responsiveness, generation speed and the wait before output begins are separate measurements. The announcement supports a reported throughput claim, without establishing how quickly an entire user task finishes.
Terms in this piece · Glossary
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.
