Realtime TTS
Engage every user with the #1 ranked, most natural Voice AI
#1 ranked TTS with human-like expression and realtime sub-200ms latency that feels like a real conversation. Custom voices with instant cloning or text-based voice design. Fully multilingual, built for streaming, and a fraction of the cost of other providers.
#1 ranked, naturally expressive voices to engage every user
Human-like expressiveness and naturalness that builds trust. Easily create custom voices with cloning and text-based design tailored to your users. Stability optimized to eliminate hallucinations and artifacts.
#1 ranked quality. Natural and expressive. Proven by real users.
3 of the top 5 models on Artificial Analysis are Inworld. Blind tests by thousands of real users, not internal evals.
#1
Ranked on
Artificial Analysis
Enhanced voice cloning in seconds
Create custom voices instantly from just 5–15 seconds of audio, remove background noise, then preview the cloned voice.
Original
Sample uploaded
Clone
Text-based voice design
Describe any voice in natural language and generate it instantly — no audio sample needed. Pick a preset to see how it works.
Voice Description
A friendly, polished female voice with a warm British accent. Clear articulation and a reassuring cadence.
Realtime latency that keeps conversations natural
Sub-200ms for Max, sub-130ms for Mini. Fast enough that users never notice. They just keep talking.
P90 Latency
<250ms P90 first chunk latency for Max, <130ms for Mini. ~4x faster than TTS-1. Even your slowest requests feel instant.
Median Latency
<200ms median first chunk latency for Max, <100ms for Mini. ~4x faster than TTS-1. Voice agents respond before users notice a delay.
Streaming-native
Built for realtime from the ground up. Audio generates the instant it's synthesized via WebSocket. No buffering delay.
Reach and resonate with any user, anywhere
Easily scale to millions of users at $0.01/min. 75% cheaper than alternatives. 15+ languages with cross-lingual cloning. On-prem deployment for control and compliance.
Enhanced multilingual
English, Spanish, French, Korean, Chinese, Hindi, Japanese, German, and more. Native-speaker quality in every language. Deploy globally without separate pipelines.
- English
- Spanish
- French
- German
- Japanese
- Korean
- Chinese
Voice AI at $0.01/min
Compare to $0.12/min with the next best option. We're $15-30 per million characters, 87% cheaper than other providers.
$0
Inworld TTS-1.5
$>0
ElevenLabs v3
On-prem deployment
Full data sovereignty for enterprises with compliance requirements. Same quality, your environment.
Full breakdown
Feature
TTS-1.5 Max
TTS-1.5 Mini
Best for
Most applications
Latency-critical applications
Pricing
$0.025/min ($30/1M characters)
$0.01/min ($15/1M characters)
P90 latency
<250ms
<130ms
Quality
Maximum expressiveness and stability
High expressiveness
Multilingual
15 languages
15 languages
Voice cloning
Professional voice cloning
Character, word, viseme and phoneme timestamps
Custom pronunciation
On-premise
Zero data retention
We recommend TTS-1.5 Max for most use cases. The enhanced stability and quality are worth the marginal latency tradeoff for the vast majority of applications.
Research
Integrations
Try Realtime TTS now
Get started with TTS-1.5 Max, the best balance of quality and speed for most applications.
FAQs
Getting started is simple. You can try Realtime TTS instantly in the TTS Playground, where you can test voices, adjust settings, and experiment with features like instant voice cloning.
When you’re ready to integrate TTS into your application, you can follow the Developer Quickstart to make your first API request in minutes. Just create an API key in the Inworld Portal, then synthesize speech with a single POST request. Inworld supports multiple output formats, including MP3, Linear PCM (WAV), and Opus, making it easy to integrate with almost any system.
Realtime TTS is evaluated through blind listening tests by thousands of real users. TTS-1.5 Max delivers over 30% more expressiveness than its predecessor, with optimized stability and natural conversational delivery.
For most applications: TTS-1.5 Max (~200ms latency, $30/1M characters)
TTS-1.5 Max offers the best balance of quality and speed. The enhanced stability means fewer edge cases, better voice cloning fidelity, and more consistent output across languages.
For latency-critical applications: TTS-1.5 Mini (<100ms latency, $15/1M characters)
Choose TTS-1.5 Mini only if minimal latency is your absolute top priority — for example, real-time gaming or ultra-responsive voice agents where every millisecond matters.
TTS-1.5 Mini achieves <120ms P90 latency. TTS-1.5 Max delivers ~200ms with enhanced stability and quality. Both support real-time streaming via WebSocket. For most applications, we recommend TTS-1.5 Max — the quality improvement is worth the marginal latency tradeoff.
Yes. Inworld provides two types of voice cloning:
Instant (zero-shot) voice cloning
- Available to all users in the Portal
- Creates a custom voice from just 15 seconds of audio
- Ready to use in minutes
Professional voice cloning
- Fine-tuned using 30+ minutes of clean audio (minimum ~5 minutes, 20+ minutes recommended for best results)
- Recommended for uncommon voice types such as children's voices or unique accents, where instant cloning may not perform well
- Currently available by contacting the Inworld sales team
Both methods allow you to use your cloned voice directly in the Playground or via API using its unique voiceId.
Inworld TTS-1.5 supports 15 languages: English, Spanish, French, Korean, Dutch, Chinese, German, Italian, Japanese, Polish, Portuguese, Russian, Hindi, Arabic, and Hebrew.
For multilingual applications, we recommend TTS-1.5 Max for the best pronunciation, intonation, and naturalness across all supported languages.
Absolutely. Realtime TTS provides several ways to customize how the speech sounds:
Voice parameters
- Temperature: Controls expressiveness and randomness
- Talking speed: 0.5× to 1.5× of the native speaking rate
Yes. Realtime TTS supports timestamp alignment for word, character, phoneme, and viseme level synchronization. This can be helpful for subtitles, captions, lipsync, and more.
You can enable it in your API request by setting timestampType to WORD or CHARACTER.
The API response includes:
- word or character tokens
- start and end timestamps (in seconds)
- structured alignment data matching the generated audio
- phoneme-level timing and viseme symbols for lip-sync (TTS 1.5 models only)
Timestamp alignment currently supports English, with other languages available experimentally. Note that enabling timestamps currently adds roughly 100 ms of additional latency.
TTS-1.5 is a major update delivering improvements across speed, quality, and accessibility:
The Fastest: <120ms P90 latency — the fastest realtime TTS available. TTS-1.5 Max delivers ~200ms with enhanced quality.
The Highest Quality: Optimized stability to minimize hallucinations, cutoffs, and artifacts. Over 30% more expressive than TTS-1.
The Most Accessible: 15 languages (including Hindi), enhanced voice cloning, on-premise H100/B200 deployment, and 25x lower cost than alternatives.
Which model should I use? For most applications, we recommend TTS-1.5 Max. Use TTS-1.5 Mini only when minimal latency is the top priority.
Yes. Inworld provides a free, open-source ElevenLabs Migration Tool that lets you batch-transfer your custom voice clones from ElevenLabs to Inworld. The tool automatically downloads your ElevenLabs voice samples, handles audio processing (format conversion, padding, and trimming), and re-clones them in Inworld.
The migration runs entirely on your local machine with direct API communication — no data is proxied through any intermediary servers. You can also preview your migrated voices with Realtime TTS before finalizing.