Voice Design API
Create a custom AI voice from a text description
Create a custom AI voice from a text description
Create a custom AI voice from a text description
Our Voice Design API is the leading AI voice generator that turns a text prompt into a brand-new, ownable voice. Serve it through the the same low-latency API powering voice at leading Fortune 500s.
10K free credits per month
Full access to all APIs
SOC 2 Type 2 and ISO 27001
Voices that sound natural and emotionally rich, backed by an API that’s solid and dependable. Made it possible to scale new languages and markets with confidence.
Voices that sound natural and emotionally rich, backed by an API that’s solid and dependable. Made it possible to scale new languages and markets with confidence.
Cankut Kostur
Sr. Engineer, Codeway
from elevenlabs import ElevenLabs client = ElevenLabs(api_key="sk_...") previews = client.text_to_voice.design( voice_description="Confident Indian male, stern but energetic", text="Your platform just found its voice.", ) # audition candidates, then save the one you want
$ curl -X POST https://api.elevenlabs.io/v1/text-to-voice \ -H "xi-api-key: $ELEVENLABS_API_KEY" \ -H "Content-Type: application/json" \ -d '{"voice_description": "Warm, mid-30s female narrator with a soft American accent", "text": "Your platform just found its voice."}'
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY, }); const previews = await client.textToVoice.design({ voiceDescription: "Bubbly, young female with a fun and dynamic Indian accent", text: "Your platform just found its voice.", });
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY! }); const previews = await client.textToVoice.design({ voiceDescription: "Warm, approachable male announcer with a positive American accent", text: "Your platform just found its voice.", }); // create the voice from a preview to use it everywhere
from elevenlabs import ElevenLabs client = ElevenLabs(api_key="sk_...") previews = client.text_to_voice.design( voice_description="Confident Indian male, stern but energetic", text="Your platform just found its voice.", ) # audition candidates, then save the one you want
$ curl -X POST https://api.elevenlabs.io/v1/text-to-voice \ -H "xi-api-key: $ELEVENLABS_API_KEY" \ -H "Content-Type: application/json" \ -d '{"voice_description": "Warm, mid-30s female narrator with a soft American accent", "text": "Your platform just found its voice."}'
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY, }); const previews = await client.textToVoice.design({ voiceDescription: "Bubbly, young female with a fun and dynamic Indian accent", text: "Your platform just found its voice.", });
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY! }); const previews = await client.textToVoice.design({ voiceDescription: "Warm, approachable male announcer with a positive American accent", text: "Your platform just found its voice.", }); // create the voice from a preview to use it everywhere
from elevenlabs import ElevenLabs client = ElevenLabs(api_key="sk_...") previews = client.text_to_voice.design( voice_description="Confident Indian male, stern but energetic", text="Your platform just found its voice.", ) # audition candidates, then save the one you want
$ curl -X POST https://api.elevenlabs.io/v1/text-to-voice \ -H "xi-api-key: $ELEVENLABS_API_KEY" \ -H "Content-Type: application/json" \ -d '{"voice_description": "Warm, mid-30s female narrator with a soft American accent", "text": "Your platform just found its voice."}'
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY, }); const previews = await client.textToVoice.design({ voiceDescription: "Bubbly, young female with a fun and dynamic Indian accent", text: "Your platform just found its voice.", });
import { ElevenLabsClient } from "elevenlabs"; const client = new ElevenLabsClient({ apiKey: process.env.ELEVENLABS_API_KEY! }); const previews = await client.textToVoice.design({ voiceDescription: "Warm, approachable male announcer with a positive American accent", text: "Your platform just found its voice.", }); // create the voice from a preview to use it everywhere
Voices that sound natural and emotionally rich, backed by an API that’s solid and dependable. Made it possible to scale new languages and markets with confidence.
Cankut Kostur
Sr. Engineer, Codeway
Trusted by engineers at the world’s leading brands
Trusted by engineers at the world’s leading brands
Get 10,000 free credits
Get 10,000 free credits
Get 10,000 free credits
Start designing your own custom voices now. Serve them across all ElevenAPI surfaces.
Start designing your own custom voices now. Serve them across all ElevenAPI surfaces.
Get 10K free credits
Sign up for free and get 10,000 credits instantly. No credit card required.
Get 10K free credits
Sign up for free and get 10,000 credits instantly. No credit card required.
Get 10K free credits
Sign up for free and get 10,000 credits instantly. No credit card required.
Grab your API key
Easily generate an API key from your dashboard. Free tier included.
Grab your API key
Easily generate an API key from your dashboard. Free tier included.
Grab your API key
Easily generate an API key from your dashboard. Free tier included.
Start building
One request returns streamed audio. Wire it into your app and you're live.
Start building
One request returns streamed audio. Wire it into your app and you're live.
Start building
One request returns streamed audio. Wire it into your app and you're live.
Your full audio stack behind a single key
Your full audio stack behind a single key
Your full audio stack behind a single key
Get your free API key
Voice Design API
Describe a voice in plain language and generate a brand-new, ownable voice in seconds. Use it across every ElevenAPI surface.
Voice Design API
Describe a voice in plain language and generate a brand-new, ownable voice in seconds. Use it across every ElevenAPI surface.
Voice Design API
Describe a voice in plain language and generate a brand-new, ownable voice in seconds. Use it across every ElevenAPI surface.
Text to Speech API
Generate lifelike speech in 70+ languages. Tune for emotional depth with v3 or ~75ms inference with Flash v2.5.
Text to Speech API
Generate lifelike speech in 70+ languages. Tune for emotional depth with v3 or ~75ms inference with Flash v2.5.
Text to Speech API
Generate lifelike speech in 70+ languages. Tune for emotional depth with v3 or ~75ms inference with Flash v2.5.
Voice Changer API
Transform audio from any voice into a target voice while preserving timing, intonation, and emotional delivery.
Voice Changer API
Transform audio from any voice into a target voice while preserving timing, intonation, and emotional delivery.
Voice Changer API
Transform audio from any voice into a target voice while preserving timing, intonation, and emotional delivery.
Music API
Generate studio-quality tracks in any genre or style from a text prompt. Royalty-free, ready for commercial use.
Music API
Generate studio-quality tracks in any genre or style from a text prompt. Royalty-free, ready for commercial use.
Music API
Generate studio-quality tracks in any genre or style from a text prompt. Royalty-free, ready for commercial use.
Sound Effects API
Generate sound effects and ambient audio from a text prompt. Every request returns four royalty-free options.
Sound Effects API
Generate sound effects and ambient audio from a text prompt. Every request returns four royalty-free options.
Sound Effects API
Generate sound effects and ambient audio from a text prompt. Every request returns four royalty-free options.
Agents API
Build production-ready voice agents with natural turn-taking, interruption handling, and conflict de-escalation.
Agents API
Build production-ready voice agents with natural turn-taking, interruption handling, and conflict de-escalation.
Agents API
Build production-ready voice agents with natural turn-taking, interruption handling, and conflict de-escalation.
Speech to Text API
Transcribe audio with perfect accuracy in 90+ languages. Word-level timestamps and speaker diarization included.
Speech to Text API
Transcribe audio with perfect accuracy in 90+ languages. Word-level timestamps and speaker diarization included.
Speech to Text API
Transcribe audio with perfect accuracy in 90+ languages. Word-level timestamps and speaker diarization included.
Voice Cloning API
Create an instant voice clone from less than a minute of audio. Use it across all ElevenLabs products and APIs.
Voice Cloning API
Create an instant voice clone from less than a minute of audio. Use it across all ElevenLabs products and APIs.
Voice Cloning API
Create an instant voice clone from less than a minute of audio. Use it across all ElevenLabs products and APIs.
Dubbing API
Translate and re-voice video content across 90+ languages. Speaker identity, timing, and tone are preserved.
Dubbing API
Translate and re-voice video content across 90+ languages. Speaker identity, timing, and tone are preserved.
Dubbing API
Translate and re-voice video content across 90+ languages. Speaker identity, timing, and tone are preserved.
Speech Engine API
Add a human-like voice layer to your chat agent in minutes. Your LLM, RAG, and architecture remain untouched.
Speech Engine API
Add a human-like voice layer to your chat agent in minutes. Your LLM, RAG, and architecture remain untouched.
Speech Engine API
Add a human-like voice layer to your chat agent in minutes. Your LLM, RAG, and architecture remain untouched.
The API Fortune 500 companies ship on
The API Fortune 500 companies ship on
The API Fortune 500 companies ship on
SSO and SCIM
SAML/OIDC, role-based access, audit logs
SSO and SCIM
SAML/OIDC, role-based access, audit logs
Custom SLAs
Reserved throughput, priority routing, support engineers on call
Data controls
Zero-retention, DPA, regional processing, no training on your data
Consent and safety tooling
Voice verification and provenance built into the cloning APIs
Pay as you go API pricing
Pay as you go API pricing
Pay as you go API pricing
The same rates regardless of your plan or volume. No hidden fees, no overages.
The same rates regardless of your plan or volume. No hidden fees, no overages.
Leading text to voice design, built for production
Leading text to voice design, built for production
Leading text to voice design, built for production
Stream as you generate
First audio chunk in ~75 ms, playback before the sentence finishes
WebSocket for live agents
Bidirectional conversational audio with sub-second turn-taking
Backpressure handled for you
Chunked transfer, auto retries, graceful degradation under load
Scale concurrency on demand
Request limits lift with your tier, no infra to manage
We are the first in the world to offer AI functions directly over the network. We are removing barriers. No apps, no special devices, no technical complexity.
We are the first in the world to offer AI functions directly over the network. We are removing barriers. No apps, no special devices, no technical complexity.
We are the first in the world to offer AI functions directly over the network. We are removing barriers. No apps, no special devices, no technical complexity.
![](data:image/svg+xml,<svg display="block" id="2851063554" role="presentation" viewBox="0 0 32 32" xmlns="http://www.w3.org/2000/svg"><path d="M 32 32 L 0 32 L 0 0 L 32 0 Z M 7.954 12.964 L 9.094 12.964 L 9.094 12.702 C 9.094 9.777 10.998 7.788 13.996 7.788 L 14.271 7.788 L 14.271 21.935 C 14.271 23.336 13.536 24.287 11.935 24.287 L 11.007 24.287 L 11.007 25.637 L 20.994 25.637 L 20.994 24.287 L 20.066 24.287 C 18.465 24.287 17.731 23.336 17.731 21.936 L 17.731 7.787 L 18.005 7.787 C 21.003 7.787 22.907 9.776 22.907 12.701 L 22.907 12.963 L 24.047 12.963 L 24.047 6.437 L 7.955 6.437 L 7.955 12.964 Z M 7.953 14.138 L 7.953 17.978 L 11.78 17.978 L 11.78 14.138 Z M 20.22 17.978 L 24.047 17.978 L 24.047 14.138 L 20.22 14.138 Z" fill="rgb(219, 2, 111)" height="32px" id="RmxBrlzs4" transform="translate(0 -0.005)" width="32px"/></svg>)
Abdu Mudesir
Product & Technology, Deutsche Telekom / T-Mobile
Made by engineers for engineers
Made by engineers for engineers
Made by engineers for engineers
ElevenLabs solves the problem of robotic, unnatural AI voiceovers. The API integration is smooth, which makes it easy to implement in applications and automation systems.
![](data:image/svg+xml,<svg display="block" id="3726906826" role="presentation" viewBox="0 0 40 40" xmlns="http://www.w3.org/2000/svg"><g fill="transparent" height="40px" id="ui5FobGT8" width="40px"><path d="M 40 20 C 40 31.046 31.046 40 20 40 C 8.955 40 0 31.046 0 20 C 0 8.955 8.955 0 20 0 C 31.046 0 40 8.955 40 20 Z" fill="rgb(255, 73, 44)" height="40px" id="rphN51G_d" width="40px"/><path d="M 8.94 5.925 L 5.522 5.925 C 5.615 5.389 5.945 5.09 6.617 4.75 L 7.247 4.43 C 8.372 3.855 8.972 3.205 8.972 2.142 C 8.972 1.474 8.714 0.949 8.197 0.568 C 7.691 0.186 7.072 0 6.359 0 C 5.791 0 5.275 0.145 4.8 0.444 C 4.335 0.731 3.984 1.103 3.767 1.566 L 4.759 2.555 C 5.14 1.782 5.697 1.401 6.431 1.401 C 7.051 1.401 7.432 1.721 7.432 2.164 C 7.432 2.535 7.247 2.844 6.534 3.205 L 6.131 3.4 C 5.254 3.844 4.644 4.349 4.294 4.925 C 3.944 5.492 3.767 6.223 3.767 7.1 L 3.767 7.336 L 8.94 7.336 Z M 8.476 8.971 L 2.826 8.971 L 0 13.866 L 5.65 13.866 L 8.475 18.761 L 11.3 13.866 L 8.475 8.973 Z" fill="rgb(255, 255, 255)" height="18.761251487731933px" id="yg4lSGsgI" transform="translate(20 9.375)" width="11.300000000000004px"/><path d="M 11.433 17.954 C 7.827 17.952 4.904 15.031 4.899 11.425 C 4.904 7.819 7.827 4.898 11.433 4.896 L 13.67 0.219 C 12.933 0.073 12.184 0 11.433 0 C 5.119 0 0 5.115 0 11.425 C 0 17.735 5.119 22.85 11.433 22.85 C 13.95 22.85 16.276 22.038 18.165 20.659 L 15.689 16.371 C 14.506 17.393 12.995 17.955 11.433 17.954 Z" fill="rgb(255, 255, 255)" height="22.850001762623435px" id="vxuVzQrYm" transform="translate(8.75 8.75)" width="18.164999999999996px"/></g></svg>)
Rexus
Verified G2 review
ElevenLabs solves the problem of robotic, unnatural AI voiceovers. The API integration is smooth, which makes it easy to implement in applications and automation systems.
Rexus
Verified G2 review
The quality is consistent even with longer text, and the API is easy to integrate. Voice customization, stability, and low latency make it a reliable tool for production use, not just demos.
Wojciech K.
Verified G2 review
The quality is consistent even with longer text, and the API is easy to integrate. Voice customization, stability, and low latency make it a reliable tool for production use, not just demos.
Wojciech K.
Verified G2 review
I like that the interface is very straightforward and the API is also easy to develop against. It's a sophisticated and varied service that only reveals the tools you want when you need them.
Adam B.
Verified G2 review
I like that the interface is very straightforward and the API is also easy to develop against. It's a sophisticated and varied service that only reveals the tools you want when you need them.
Adam B.
Verified G2 review
Another new thing I like is the enhanced API flexibility, which allows developers to integrate ElevenLabs into a wider range of applications — from podcasts and audiobooks to real-time conversational agents — without losing quality or responsiveness.
Sonu K.
Verified G2 review
Another new thing I like is the enhanced API flexibility, which allows developers to integrate ElevenLabs into a wider range of applications — from podcasts and audiobooks to real-time conversational agents — without losing quality or responsiveness.
Sonu K.
Verified G2 review
The initial setup was very simple, and I use the API extensively in my projects, including English learning apps and story apps since I am an Android and iOS developer.
Anand B.
Verified G2 review
The initial setup was very simple, and I use the API extensively in my projects, including English learning apps and story apps since I am an Android and iOS developer.
Anand B.
Verified G2 review
The API's ability to automate speech-to-text conversions for large batches in multiple languages, including Chinese and Korean, directly enhances our workflow.
Vedaswaroop I.
Verified G2 review
The API's ability to automate speech-to-text conversions for large batches in multiple languages, including Chinese and Korean, directly enhances our workflow.
Vedaswaroop I.
Verified G2 review
Initial setup of ElevenLabs is impressively straightforward, allowing for quick access to API integrations and generating voiceovers efficiently.
Mohdhafiz L.
Verified G2 review
I enjoy how seamlessly ElevenLabs integrates into various workflows through its web interface, API, and app integrations, making it a versatile solution for both casual and professional use.
Chinonso N.
Verified G2 review
The voices sound incredibly realistic, and there are plenty of options to choose from. I appreciate being able to create a voice from text, and the API responses are impressively fast.
Erick C.
Verified G2 review
The entire setup process of ElevenLabs was very simple and quick, including finding the API key, which has significantly facilitated my workflow.
Lucello H.
Verified G2 review
I really like how natural and human-like the voices sound. The Text-to-Speech API is easy to integrate with Python, and the response time is very fast. It works perfectly with my AI assistant project.
Agha E.
Verified G2 review
The text-to-speech capability is truly magical, significantly enhancing my experience when reading scholarly articles and using the API as an answering assistant for my phone.
Ajibola L.
Verified G2 review
The initial setup was very easy. From creating the account to making the first successful API request only took about 20 minutes.
Scott H.
Verified G2 review
Documentation is excellent, making setup straightforward. Having just to grab an API key simplifies the process significantly. The cost-effectiveness is a huge plus as well.
Livan C.
Verified G2 review
As a software engineer working at an AI startup, my experience with ElevenLabs has been overwhelmingly positive. The dashboard is intuitive, and their API documentation is clear and beginner-friendly.
Esha D.
Verified G2 review
I like how easy the integration and overall ease of use are. I value the voice cloning API and usage dashboard a lot. It's particularly impressive how easy it is to add emotions and clone voices.
Jaydeep R.
Verified G2 review
It’s been essential in automating voice responses in our app, allowing us to convert AI-generated text into voice and deliver it at scale with zero friction. The API is fast, reliable, and plays nicely with other tools, which has been a massive productivity boost.
Lee H.
Verified G2 review
I use ElevenLabs to power my AI storytelling app and the quality of the voices truly brings the experience to life. Integration was fast thanks to the simple API. Users frequently comment on how "human" the narration feels, which is exactly what I wanted.
Kevin E.
Verified G2 review
Answers to common questions
Answers to common questions
Answers to common questions
Yes - that's Voice Design, not cloning. Describe the voice you want in plain language (age, accent, gender, tone, character) and the API generates brand-new voices from your prompt alone, with no sample audio or real person required. Cloning copies a voice that already exists; Voice Design invents one. Save the candidate you like and serve it across every ElevenLabs API, the same as any other voice.
The REST API has official SDKs in Python and TypeScript, with documentation covering streaming TTS, voice cloning, real-time STT, and full ElevenAgents setups including tool use and webhooks. Most developers make their first API call within minutes of signing up.
Flash v2.5 delivers ~75ms model inference time, making it well-suited for low-latency voice applications. For real-time transcription, Scribe v2 Realtime transcribes in under 150ms. For use cases where expressiveness or language breadth matters more than speed, Eleven v3 and Multilingual v2 are available.
ElevenLabs holds SOC 2 Type 2, ISO 27001, ISO 27018, HIPAA (attestation), GDPR (audit), and PCI DSS Level 1 certifications, among others. Data is encrypted in transit and at rest. For stricter data requirements, regional data residency and Zero Retention Mode are available.
Eleven v3 supports 70+ languages. Flash v2.5 - optimized for low latency - supports 32 languages. Dubbing v2 supports 90+ languages. You can choose from 10,000+ pre-built voices in the Voice Library or create your own.
Instant voice cloning works with less than a minute of sample audio. For branded voices or distinctive vocal characteristics - accent, tone, delivery style - Professional Voice Cloning captures deeper nuance. Both are available through the API so you can clone and serve voices programmatically at scale.
ElevenLabs offers a free plan to explore the API. Paid plans scale by usage, with options for individual developers, growing teams, and enterprise. Enterprise includes custom rate limits, SLAs, and dedicated support.
It's a full platform. ElevenAgents handles the full conversational loop - turn detection, interruption handling, tool calls, and real-time voice - without requiring you to stitch together separate ASR, TTS, and orchestration providers. You can configure, deploy, and monitor agents from a single platform.
Voice quality is the reason most developers evaluate ElevenLabs in the first place. The models produce AI voices that sound virtually indistinguishable from human recordings - with accurate prosody, natural pacing, and emotional range across accents and languages. The best way to verify this for your use case is to run audio through the API directly; no sales call required.
Our docs cover the full API surface - SDKs, streaming, webhooks, ElevenAgents configuration, and Scribe. There is an active developer community on Discord for peer support and feedback. Paid plans include email support, and enterprise plans include dedicated support with defined SLAs.
Yes - that's Voice Design, not cloning. Describe the voice you want in plain language (age, accent, gender, tone, character) and the API generates brand-new voices from your prompt alone, with no sample audio or real person required. Cloning copies a voice that already exists; Voice Design invents one. Save the candidate you like and serve it across every ElevenLabs API, the same as any other voice.
The REST API has official SDKs in Python and TypeScript, with documentation covering streaming TTS, voice cloning, real-time STT, and full ElevenAgents setups including tool use and webhooks. Most developers make their first API call within minutes of signing up.
Flash v2.5 delivers ~75ms model inference time, making it well-suited for low-latency voice applications. For real-time transcription, Scribe v2 Realtime transcribes in under 150ms. For use cases where expressiveness or language breadth matters more than speed, Eleven v3 and Multilingual v2 are available.
ElevenLabs holds SOC 2 Type 2, ISO 27001, ISO 27018, HIPAA (attestation), GDPR (audit), and PCI DSS Level 1 certifications, among others. Data is encrypted in transit and at rest. For stricter data requirements, regional data residency and Zero Retention Mode are available.
Eleven v3 supports 70+ languages. Flash v2.5 - optimized for low latency - supports 32 languages. Dubbing v2 supports 90+ languages. You can choose from 10,000+ pre-built voices in the Voice Library or create your own.
Instant voice cloning works with less than a minute of sample audio. For branded voices or distinctive vocal characteristics - accent, tone, delivery style - Professional Voice Cloning captures deeper nuance. Both are available through the API so you can clone and serve voices programmatically at scale.
ElevenLabs offers a free plan to explore the API. Paid plans scale by usage, with options for individual developers, growing teams, and enterprise. Enterprise includes custom rate limits, SLAs, and dedicated support.
It's a full platform. ElevenAgents handles the full conversational loop - turn detection, interruption handling, tool calls, and real-time voice - without requiring you to stitch together separate ASR, TTS, and orchestration providers. You can configure, deploy, and monitor agents from a single platform.
Voice quality is the reason most developers evaluate ElevenLabs in the first place. The models produce AI voices that sound virtually indistinguishable from human recordings - with accurate prosody, natural pacing, and emotional range across accents and languages. The best way to verify this for your use case is to run audio through the API directly; no sales call required.
Our docs cover the full API surface - SDKs, streaming, webhooks, ElevenAgents configuration, and Scribe. There is an active developer community on Discord for peer support and feedback. Paid plans include email support, and enterprise plans include dedicated support with defined SLAs.
Ship something people can hear
Ship something people can hear
Ship something people can hear
Get 10,000 free credits, start building now.
Get 10,000 free credits, start building now.