Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Inworld AI / TTS-1.5

Realtime TTS-1.5 Max

Previous-generation speech model retained for existing integrations, with 15-language coverage and temperature-based delivery controls.

Speech & voice
deprecatedClosed weightsText-to-speech
Official documentation (opens in a new tab)Reviewed 13 Oct 2026
12 of 22 specification fields collected
Text input limitUnary 2,000 UTF-16 code units; HTTP stream 4,000; WebSocket 2,000 per sendText messageSee specification notes below
Languages15 listedFull language list in specifications
Input billingNot collectedInworld API
Output billingNot collectedNo inferred price

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
Not collected
Maximum output
Not collected
Input modalities
TextSource for Input modalities (opens in a new tab)
Output modalities
AudioSource for Output modalities (opens in a new tab)
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected
Streaming
SupportedSource for Streaming (opens in a new tab)
Batch processing
Not supportedSource for Batch processing (opens in a new tab)

The current native batch endpoint accepts TTS-2 only.

Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2026-01-21Source for Release date (opens in a new tab)
Parameters
Not collected
Architecture
Not collected
Languages
View all 15 entries
  • Arabic (ar)
  • Chinese (zh)
  • Dutch (nl)
  • English (en)
  • French (fr)
  • German (de)
  • Hebrew (he)
  • Hindi (hi)
  • Italian (it)
  • Japanese (ja)
  • Korean (ko)
  • Polish (pl)
  • Portuguese (pt)
  • Russian (ru)
  • Spanish (es)
Source for Languages (opens in a new tab)

The current host table lists these 15 languages for both 1.5 variants; the native release notes also establish a 15-language count.

Context caching
Not collected
Audio controls
Temperature-based variation and speaking-rate controls; TTS-2 steering/deliveryMode are not supportedSource for Audio controls (opens in a new tab)
Maximum prompt characters
Unary 2,000 UTF-16 code units; HTTP stream 4,000; WebSocket 2,000 per sendText messageSource for Maximum prompt characters (opens in a new tab)

Endpoint-specific limits; longer async jobs are available only on TTS-2. OpenAI-compatible input is documented separately as up to 4,000 characters.

Output formats
MP3, PCM, WAV, LINEAR16, OGG_OPUS, FLAC, ALAW, MULAWSource for Output formats (opens in a new tab)

Encoding support and sample rate vary by transport; OpenAI-compatible synthesis supports only its documented five formats.

Audio sample rates
Requested output 8,000–48,000 Hz; default 48,000 Hz with codec-specific restrictionsSource for Audio sample rates (opens in a new tab)

Resampling support is not a native generation-rate claim.

Voice cloning
Instant cloning from as little as three seconds of reference audio; account voice IDs are reusable across supported modelsSource for Voice cloning (opens in a new tab)
Supplied context limit
Up to 2,000 characters of earlier synthesis text; not billedSource for Supplied context limit (opens in a new tab)

synthesisContext supplies prior request text. It is not a model token window or a prompt-cache guarantee.

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

Inworld API

Provider source (opens in a new tab)
API model ID
inworld-tts-1.5-max
Access
Deprecated model; confirm entitlement and native rate for existing integrations.
Regions
General API region not pinned by these contracts. EU and India residency are enterprise offerings, not assumed for every request.
API endpoint URL
https://api.inworld.ai/tts/v1/voiceSource for API endpoint URL (opens in a new tab)
Streaming
Not supportedSource for Streaming (opens in a new tab)
Maximum prompt characters
2,000 UTF-16 code unitsSource for Maximum prompt characters (opens in a new tab)
Request mode
Single JSON object with Base64 audioSource for Request mode (opens in a new tab)
Request quotas
TTS generation concurrency by plan: 5 / 10 / 50 / 150 / 500. WebSocket connections: 50 / 100 / 500 / 1,500 / 5,000; enterprise customSource for Request quotas (opens in a new tab)

Account-level limits are shared across keys. Active generations are distinct from idle WebSocket connections.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Inworld terms exclude nonpublic inputs/outputs from general model training, with feedback and trust/safety exceptions. ZDR is configurable and has product-specific exceptions; it is not a universal default. Pricing lists a Growth add-on, while the ZDR guide limits self-serve controls to Enterprise/Enterprise Trial. Confirm the actual workspace setting and contract. Clone samples and voice-design creation inputs are retained.

Read the service policy (opens in a new tab)

Inworld API

Provider source (opens in a new tab)
API model ID
inworld-tts-1.5-max
Access
Deprecated model; confirm entitlement and native rate for existing integrations.
Regions
General API region not pinned by these contracts. EU and India residency are enterprise offerings, not assumed for every request.
API endpoint URL
https://api.inworld.ai/tts/v1/voice:streamSource for API endpoint URL (opens in a new tab)
Streaming
SupportedSource for Streaming (opens in a new tab)
Maximum prompt characters
4,000 UTF-16 code unitsSource for Maximum prompt characters (opens in a new tab)
Request mode
Streaming JSON results with Base64 audio chunksSource for Request mode (opens in a new tab)
Request quotas
TTS generation concurrency by plan: 5 / 10 / 50 / 150 / 500. WebSocket connections: 50 / 100 / 500 / 1,500 / 5,000; enterprise customSource for Request quotas (opens in a new tab)

Account-level limits are shared across keys. Active generations are distinct from idle WebSocket connections.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Inworld terms exclude nonpublic inputs/outputs from general model training, with feedback and trust/safety exceptions. ZDR is configurable and has product-specific exceptions; it is not a universal default. Pricing lists a Growth add-on, while the ZDR guide limits self-serve controls to Enterprise/Enterprise Trial. Confirm the actual workspace setting and contract. Clone samples and voice-design creation inputs are retained.

Read the service policy (opens in a new tab)

Inworld API

Provider source (opens in a new tab)
API model ID
inworld-tts-1.5-max
Access
Deprecated model; confirm entitlement and native rate for existing integrations.
Regions
General API region not pinned by these contracts. EU and India residency are enterprise offerings, not assumed for every request.
API endpoint URL
wss://api.inworld.ai/tts/v1/voice:streamBidirectionalSource for API endpoint URL (opens in a new tab)
Streaming
SupportedSource for Streaming (opens in a new tab)
Maximum prompt characters
2,000 UTF-16 code units per sendText messageSource for Maximum prompt characters (opens in a new tab)
Request mode
Bidirectional text/audio contexts; explicit flush and close ; use CLIENT_SEGMENTED, not TTS-2-only sentence-boundary modeSource for Request mode (opens in a new tab)
Request quotas
TTS generation concurrency by plan: 5 / 10 / 50 / 150 / 500. WebSocket connections: 50 / 100 / 500 / 1,500 / 5,000; enterprise customSource for Request quotas (opens in a new tab)

Account-level limits are shared across keys. Active generations are distinct from idle WebSocket connections.

Session duration
Session closes after ten minutes of inactivity across contextsSource for Session duration (opens in a new tab)

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Inworld terms exclude nonpublic inputs/outputs from general model training, with feedback and trust/safety exceptions. ZDR is configurable and has product-specific exceptions; it is not a universal default. Pricing lists a Growth add-on, while the ZDR guide limits self-serve controls to Enterprise/Enterprise Trial. Confirm the actual workspace setting and contract. Clone samples and voice-design creation inputs are retained.

Read the service policy (opens in a new tab)

Inworld API

Provider source (opens in a new tab)
API model ID
inworld-tts-1.5-max
Access
Deprecated model; confirm entitlement and native rate for existing integrations.
Regions
General API region not pinned by these contracts. EU and India residency are enterprise offerings, not assumed for every request.
API endpoint URL
https://api.inworld.ai/v1/audio/speechSource for API endpoint URL (opens in a new tab)
Streaming
SupportedSource for Streaming (opens in a new tab)
Maximum prompt characters
4,000 charactersSource for Maximum prompt characters (opens in a new tab)
Request mode
Raw audio bytes or Server-Sent EventsSource for Request mode (opens in a new tab)
Request quotas
TTS generation concurrency by plan: 5 / 10 / 50 / 150 / 500. WebSocket connections: 50 / 100 / 500 / 1,500 / 5,000; enterprise customSource for Request quotas (opens in a new tab)

Account-level limits are shared across keys. Active generations are distinct from idle WebSocket connections.

Output formats
mp3, opus, flac, wav, pcmSource for Output formats (opens in a new tab)
Audio sample rates
MP3/Opus/FLAC/WAV: 48 kHz mono; raw PCM16: 24 kHz monoSource for Audio sample rates (opens in a new tab)
Authentication
Bearer API key; other native TTS routes use Basic credentialsSource for Authentication (opens in a new tab)

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Inworld terms exclude nonpublic inputs/outputs from general model training, with feedback and trust/safety exceptions. ZDR is configurable and has product-specific exceptions; it is not a universal default. Pricing lists a Growth add-on, while the ZDR guide limits self-serve controls to Enterprise/Enterprise Trial. Confirm the actual workspace setting and contract. Clone samples and voice-design creation inputs are retained.

Read the service policy (opens in a new tab)

LiveKit Inference

Provider source (opens in a new tab)
API model ID
inworld/inworld-tts-1.5-max
Access
Listed in LiveKit Inference; no separate Inworld key required. Host plan, default voice support and account limits apply.
Regions
Global. The model row does not list a dedicated EU endpoint.
Languages
View all 15 entries
  • en
  • zh
  • ja
  • ko
  • ru
  • it
  • es
  • pt
  • fr
  • de
  • pl
  • nl
  • hi
  • he
  • ar
Source for Languages (opens in a new tab)

Host language table is retained independently of native tiers and locale aliases.

Streaming
SupportedSource for Streaming (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 13 Oct 2026
CategoryUsageRateUnit
Speech synthesisSpeech synthesis · Build / Ship1$35.001M characters
Speech synthesisSpeech synthesis · Scale1$20.001M characters

Rate conditions

  1. Published LiveKit Inference rate; plan fees, included credits, default-voice compatibility and enterprise terms are separate.
Pricing source (opens in a new tab)

Data & enterprise terms

LiveKit documents zero retention for inference content by default, including underlying providers. Optional observability and custom-voice storage have separate handling.

Read the service policy (opens in a new tab)

Cloudflare AI

Provider source (opens in a new tab)
API model ID
inworld/tts-1.5-max
Access
Third-party model served through Cloudflare AI. Follow this host’s schema and account access rules; the native deprecated status does not prove host retirement.
Regions
Cloudflare deployment; exact data residency for this model uncollected.
Maximum prompt characters
2,000 charactersSource for Maximum prompt characters (opens in a new tab)
Generation controls
Speaking rate 0.5–1.5; temperature >0 to 2; timestamp none/word/characterSource for Generation controls (opens in a new tab)
Request mode
Host response exposes an audio URL, not the native Base64 response contractSource for Request mode (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 13 Oct 2026
CategoryUsageRateUnit
Speech synthesisSpeech synthesis1$0.000035character

Rate conditions

  1. Published Cloudflare rate. Native and LiveKit subscription rates are not applied here.
Pricing source (opens in a new tab)

Data & enterprise terms

Host-specific retention/training terms were not established from this model card. Do not inherit native Inworld or LiveKit guarantees.

Read the service policy (opens in a new tab)

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

Inworld API

inworld-tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Inworld API

inworld-tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Inworld API

inworld-tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Inworld API

inworld-tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

LiveKit Inference

inworld/inworld-tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Cloudflare AI

inworld/tts-1.5-max
Not verified

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How it works
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Minimum reusable input
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Lifetime & refresh
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

How to enable it
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Usage fields
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Billing rules
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

Limits & data boundaries
Not collected

No model-specific prompt-cache mechanism, lifetime or cached-token price established. Free synthesisContext text and reusable voice profiles do not establish prompt caching.

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Voice choices

3 voices

Choose a voice within Realtime TTS-1.5 Max. These are API voice identifiers, not additional model releases. Source links document each voice.

Published voice catalog
Voice / API IDLanguage & accentDelivery
AlexAlexEnglishEnergetic and expressive mid-range male voice, with a mildly nasal quality. Public catalog example, not the full workspace voice inventory.Voice source (opens in a new tab)
AshleyAshleyEnglishA warm, natural female voice. Public catalog example, not the full workspace voice inventory.Voice source (opens in a new tab)
DennisDennisEnglishMiddle-aged man with a smooth, calm and friendly voice. Public catalog example, not the full workspace voice inventory.Voice source (opens in a new tab)

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Inworld grants limited service/model access and reserves model IP under its proprietary terms, including separately contracted customer-hosted deployment. The TTS-1 research release publishes training/modeling code, not the production TTS model weights. Current TTS/STT release identities are established in the native catalog; routed third-party models are excluded.

Classification evidence (opens in a new tab)

Lifecycle dates

Native model release21 Jan 2026Source (opens in a new tab)

Sources reviewed