All models
Voice & realtimeStable
Inworld AI
Inworld Realtime TTS-2
Inworld's newest expressive realtime text-to-speech model, designed to remember conversational delivery and speak in more than 100 languages.
Plain-English overview
What Inworld Realtime TTS-2 actually is
Inworld Realtime TTS-2 turns text into streamed speech while using earlier audio turns to keep a character's delivery consistent. Developers can describe the desired mood or delivery in natural language, switch languages, and use voice cloning where their rights and account setup permit it.
Pricing is based on characters rather than audio tokens, which makes scripts easier to estimate. Inworld reports sub-200ms median time to first audio, but network distance, application work, and the upstream language model still affect the complete response time.
Good fit for
- Realtime companions and support voices
- Multilingual experiences with one consistent voice
- Creative voice direction and authorized cloning
Category comparison
The facts that matter for voice models
These are provider-published specifications, not Cody benchmark scores. Follow the linked sources for current limits and endpoint-specific exceptions.
- Price basisToken-, character-, or minute-based price for the listed access route.
- $25 / 1M chars on demand; volume discounts
- Interaction modeSpeech-to-speech, audio-to-audio, or streamed text-to-speech behavior.
- Steerable realtime TTS with prior-audio context
- Published latencyProvider-published measurement with its stated exclusions; not a Cody test.
- Under 200ms median TTFA
- LanguagesProvider-documented language coverage or a clear unpublished marker.
- 100+ with mid-utterance switching
- Tools & controlNative function calling, reasoning, or speech-control features.
- Voice direction, design, cloning, non-verbals
- Context or input limitDocumented token or character limit where it is meaningful.
- Not published
API and provider access
Where to get Inworld Realtime TTS-2
Availability
Regions and access stage
Generally available through Inworld's REST TTS API and OpenAI-compatible Realtime API.
Standard region coverage is not listed on the model page; enterprise plans advertise EU and India data-residency options.
Check live availabilityData and training
The route matters.
Inworld advertises zero data retention as an enterprise add-on, but its currently published ZDR support page names only TTS-1.5 Mini and Max. Confirm TTS-2 coverage directly before sending regulated or confidential text.
This is a concise reading of the cited provider material, not legal advice. A third-party gateway can have different storage, routing, training, and residency terms from the model maker's direct API.
Read the provider policyFrequently asked questions
Inworld Realtime TTS-2 FAQ
What is Inworld Realtime TTS-2?
Inworld's newest expressive realtime text-to-speech model, designed to remember conversational delivery and speak in more than 100 languages. Inworld Realtime TTS-2 turns text into streamed speech while using earlier audio turns to keep a character's delivery consistent. Developers can describe the desired mood or delivery in natural language, switch languages, and use voice cloning where their rights and account setup permit it.
When was Inworld Realtime TTS-2 released?
Inworld Realtime TTS-2 was released on August 31, 2026 according to the cited provider materials.
Where can I access Inworld Realtime TTS-2?
Generally available through Inworld's REST TTS API and OpenAI-compatible Realtime API. The access routes listed in this guide are Inworld AI, Inworld AI, Inworld AI.
How much does Inworld Realtime TTS-2 cost?
$25 / 1M characters on demand. Published subscription rates fall to $20, $17.50, $15, and $12.50 per million characters by tier, with enterprise pricing advertised as low as $5.
Where is Inworld Realtime TTS-2 available?
Generally available through Inworld's REST TTS API and OpenAI-compatible Realtime API. Standard region coverage is not listed on the model page; enterprise plans advertise EU and India data-residency options.
Is my Inworld Realtime TTS-2 API data used for training?
Inworld advertises zero data retention as an enterprise add-on, but its currently published ZDR support page names only TTS-1.5 Mini and Max. Confirm TTS-2 coverage directly before sending regulated or confidential text. The policy belongs to the provider route and account terms, so verify it again before production use.
Research trail
Sources for this guide
Keep comparing
More voice & realtime models
OpenAI · Stable
GPT-Realtime-2.1
OpenAI's realtime speech-to-speech reasoning model for tool-using voice agents that also need text and image context.
Read guideGoogle · Preview
Gemini 3.1 Flash Live Preview
Google's preview audio-to-audio model for low-latency dialogue with multimodal awareness, thinking, search grounding, and function calling.
Read guideElevenLabs · Stable
Eleven v3 Conversational
ElevenLabs' expressive realtime text-to-speech model for natural dialogue, emotional delivery, audio tags, and more than 70 languages.
Read guide