Eleven v3 is a premium text to speech model for production audio. It supports 70+ languages with studio grade quality and precise expressive control using inline audio tags. Ideal for narration, podcasts, dialogue, audiobooks, and game voiceover where stable prosody matters.

Eleven Flash v2.5 is a real time text to speech model for voice agents and interactive apps. It delivers natural speech in about 75 ms latency across 32 languages. Use it for low latency conversational AI, games, live tools, and large scale TTS workloads.

Eleven Multilingual v2 is a high fidelity multilingual text to speech model for 29 languages. It supports expressive prosody with emotional nuance. Ideal for audiobooks, localization pipelines, customer support and international applications that require natural neural voices.

Eleven Turbo v2.5 delivers fast text to speech for production apps. It targets low latency flows with rich voice quality in 32 languages. Use it to power interactive agents, games, and voice enabled tools that need natural speech with rapid response.

Eleven Flash v2 is an earlier English speech model that delivers very low latency and clear audio. It is built for live streaming use cases. It also fits real time gaming and interactive tools where rapid voice feedback is critical.