ElevenLabs' most expressive, human-like text-to-speech model supporting 70+ languages. Generates speech via the /v1/audio/speech endpoint.
eleven-v3
价格按百万 tokens 计，按 token 用量计费。