ElevenLabs speech and music models are 50% off on Vercel AI Gateway

Vercel and ElevenLabs have brought 14 ElevenLabs speech and music models to Vercel AI Gateway. To celebrate the launch, every ElevenLabs model on AI Gateway is 50% off for the first 14 days, through November 9. If you want to add transcription, voiceovers, a realtime voice assistant, or generated music to your app, this is a good window to try several of them.

What you get

Through November 9, all 14 ElevenLabs models on Vercel AI Gateway are half price. Eleven v4, the multi-voice dialogue model, costs $40 per million characters instead of $80. AI Gateway passes provider pricing through with no markup and no platform fee on inference, so the discounted rate is what you pay. Any Vercel user with AI Gateway access gets it automatically, with no coupon or promo code needed.

Vercel users who haven't made a payment yet get $5 of AI Gateway credits every 30 days, which is enough to try the models in the playground first.

What you can use

The launch covers six kinds of audio models, all reachable with one AI Gateway API key:

• Speech to text: elevenlabs/scribe-v2 and elevenlabs/scribe-v2-realtime for recordings and live captions.
• Streaming text to speech: elevenlabs/eleven-flash-v2.5, eleven-flash-v2, eleven-multilingual-v2, and eleven-v3-conversational, which start speaking while the language model is still writing.
• Realtime voice agents: elevenlabs/eleven-realtime-v1-max and eleven-realtime-v1-mini handle listening and speaking in one WebSocket session, a good fit for support lines, tutors, and in-app assistants.
• Music: elevenlabs/music-v2.5 and music-v2 turn a text description into a finished song or instrumental at the length you choose.
• Sound effects: elevenlabs/sound-effects-v2 for game and interface sounds or seamlessly looping ambience.
• Multi-voice dialogue: elevenlabs/eleven-v4, eleven-v4-turbo, and eleven-v3 voice a whole conversation in one request, with a different voice for each speaker.

How to use

1. Sign in to Vercel and create an AI Gateway API key.
2. Call the models through the AI SDK using the elevenlabs/ model IDs above. Speech and music use generateSpeech, and transcription uses streamTranscribe.
3. To listen first, open any model in the ElevenLabs model list on Vercel and try it in the playground.

import { generateSpeech } from 'ai';

const { audio } = await generateSpeech({
  model: 'elevenlabs/music-v2.5',
  text: 'A mellow lo-fi beat with soft piano',
  providerOptions: { elevenlabs: { musicLengthMs: 30000 } },
});

See the Vercel launch post and the AI Gateway docs.