Minimum one character. This is a character limit, not a native token context window. Compose shares the same 2,000-character budget across all speech segments.
Pitch -12 to +12 semitones; tempo 0.5–2.0. Standard/timestamped output volume 0–200 or target_lufs -70–0, mutually exclusive. Streaming omits volume and supports target_lufs. Values describe audio controls, not model reasoning or fine-tuning.Source for Advanced controls (opens in a new tab)
Source audio handling
remove_silence_ms is an integer 0–1,000 specifying silence to retain. Zero enables removal; null/omission disables it. Returned timestamps follow processed audio; explicit Compose pause segments are preserved.Source for Source audio handling (opens in a new tab)
Response format
Standard and streaming return binary audio. /with-timestamps returns base64 audio and word/character timing arrays; granularity=word or char selects an array. Japanese/Chinese without whitespace may form a single word segment, so request character alignment. Fixed timing JSON is not arbitrary structured generation.Source for Response format (opens in a new tab)
Bulk processing
Compose accepts up to 50 ordered speech/pause segments, at least one speech segment, 2,000 total text characters, pauses up to 10 seconds each and 60 seconds total, and a shared audio format. Voices/models may differ by segment. Segments run in parallel and return one ordered audio file; any segment failure fails the request without credits. Pauses are free.Source for Bulk processing (opens in a new tab)
Stock voice catalog
GET /v3/voices returns built-ins (tc_) and completed owner custom voices (uc_), localized names, per-model emotions and optional previews. Filter by model, voice type, age, gender or use case. Marketing pages say 600+ API voices and 700+ Studio voices; these are not a verified production API inventory.Source for Stock voice catalog (opens in a new tab)
Do not convert the 12 public sample names or example response IDs into a full model-specific voice catalog. Account access is required for authoritative availability.
Voice cloning
Instant and Professional cloning accept the chosen SSFM model, subject to account/release availability. Custom voices require rights and consent, remain on provider infrastructure, and may be used only by the owner during their subscription.Source for Voice cloning (opens in a new tab)
Voice reference requirements
Instant: one WAV/MP3, <=25 MiB, 5–150 seconds; HTTP 201 returns a completed voice. Professional: one WAV/MP3, <=1 GiB, 5 minutes–3 hours, >=16 kHz, ISO-639-3 language; HTTP 202 starts training. Both names are <=30 characters. Professional training may take up to two hours; poll custom-voice status. These are enrollment limits, not synthesis input limits.Source for Voice reference requirements (opens in a new tab)
Instant limits are separately documented in the linked instant-clone source. The legacy clone route is deprecated.
Custom voice quota
Lite: 50 total slots shared across clone types, at most 2 Professional slots and 5 Professional creations/month. Plus: 800 total slots, at most 5 Professional slots and 15 creations/month. Enterprise terms are custom. Current account capacity comes from /v1/users/me/subscription.Source for Custom voice quota (opens in a new tab)
Task statuses
Custom voice statuses: pending, training, completed, failed. A 202 training response is not a ready voice. Only completed custom voices appear in the synthesis catalog.Source for Task statuses (opens in a new tab)
Published price
USD plans: Free includes 15,000 credits/month and concurrency 2; Lite $15/month includes 200,000 credits and concurrency 5; Plus $280/month includes 4M credits and concurrency 15. One text character consumes one credit. Lite overage is $0.09/1,000 credits, charged in $9 blocks of 100,000; Plus is $0.08/1,000, charged in $100 blocks of 1.25M. Included-credit unit averages ($0.075/$0.07 per 1,000) require the full subscription. These are character charges, not token rates. Studio plans are separate.Source for Published price (opens in a new tab)
Request quotas
Concurrency limits: Free 2, Lite 5, Plus 15. These are concurrent requests, not requests per second. HTTP 429 may indicate a rate or concurrency limit.Source for Request quotas (opens in a new tab)
Model terms
Free-generated output is noncommercial and needs Typecast attribution. Paid output rights remain subject to API terms. API and Studio subscriptions are independent.Source for Model terms (opens in a new tab)
Creation privacy
API terms allow use of content to provide and improve services and explicitly authorize training/improvement using consented cloning recordings, with opt-out and deletion requests available. Custom voice models remain on provider infrastructure. The privacy policy retains voice generation records until membership withdrawal, with separate statutory retention for transactions and access logs. No default no-training or zero-retention guarantee is established; exact inference region is not published in the reviewed material.Source for Creation privacy (opens in a new tab)
Documentation discrepancy
The marketing timeline labels SSFM 3.0 as 2025; the API changelog dates v30 to January 2026. The voice filter description mentions ssfm-v31, but the enum and release directory only list v30/v21. Some subscription examples and links still reference older slot counts and deprecated clone routes; use current pricing and custom-voices contracts.Source for Documentation discrepancy (opens in a new tab)
Access & pricing
Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.
POST https://api.typecast.ai/v1/text-to-speech. X-API-KEY required. Complete binary audio response. WAV at 44.1 kHz; MP3 at 44.1 kHz/320 kbps. Use a voice compatible with the selected model.
Regions
Exact inference region not established. Korean operator location and U.S. analytics subprocessors do not establish model residency.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 16 Oct 2026
Category
Usage
Rate
Unit
API synthesis credits
Synthesized text1
1 credit
character
Lite overage
Additional synthesized text2
$0.09
1,000 characters
Plus overage
Additional synthesized text3
$0.08
1,000 characters
Rate conditions
One credit per text character. Monthly plan allowance and overage blocks apply. Compose bills combined text and no pause credits; a failed composition consumes no credits. Smart-context field billing details beyond the synthesized text meter are not independently verified.
Lite paid subscription required. Published overage unit rate; actual billing occurs in $9 per 100,000-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
Plus paid subscription required. Published overage unit rate; actual billing occurs in $100 per 1.25M-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
API terms allow use of content to provide and improve services and explicitly authorize training/improvement using consented cloning recordings, with opt-out and deletion requests available. Custom voice models remain on provider infrastructure. The privacy policy retains voice generation records until membership withdrawal, with separate statutory retention for transactions and access logs. No default no-training or zero-retention guarantee is established; exact inference region is not published in the reviewed material.
POST https://api.typecast.ai/v1/text-to-speech/stream. X-API-KEY required. Chunked response. WAV is 32 kHz PCM; MP3 is 44.1 kHz/320 kbps. Volume is not supported in streaming settings. Use a voice compatible with the selected model.
Regions
Exact inference region not established. Korean operator location and U.S. analytics subprocessors do not establish model residency.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 16 Oct 2026
Category
Usage
Rate
Unit
API synthesis credits
Synthesized text1
1 credit
character
Lite overage
Additional synthesized text2
$0.09
1,000 characters
Plus overage
Additional synthesized text3
$0.08
1,000 characters
Rate conditions
One credit per text character. Monthly plan allowance and overage blocks apply. Compose bills combined text and no pause credits; a failed composition consumes no credits. Smart-context field billing details beyond the synthesized text meter are not independently verified.
Lite paid subscription required. Published overage unit rate; actual billing occurs in $9 per 100,000-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
Plus paid subscription required. Published overage unit rate; actual billing occurs in $100 per 1.25M-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
API terms allow use of content to provide and improve services and explicitly authorize training/improvement using consented cloning recordings, with opt-out and deletion requests available. Custom voice models remain on provider infrastructure. The privacy policy retains voice generation records until membership withdrawal, with separate statutory retention for transactions and access logs. No default no-training or zero-retention guarantee is established; exact inference region is not published in the reviewed material.
POST https://api.typecast.ai/v1/text-to-speech/with-timestamps. X-API-KEY required. JSON containing base64 audio and optional word/character alignments. Use a voice compatible with the selected model.
Regions
Exact inference region not established. Korean operator location and U.S. analytics subprocessors do not establish model residency.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 16 Oct 2026
Category
Usage
Rate
Unit
API synthesis credits
Synthesized text1
1 credit
character
Lite overage
Additional synthesized text2
$0.09
1,000 characters
Plus overage
Additional synthesized text3
$0.08
1,000 characters
Rate conditions
One credit per text character. Monthly plan allowance and overage blocks apply. Compose bills combined text and no pause credits; a failed composition consumes no credits. Smart-context field billing details beyond the synthesized text meter are not independently verified.
Lite paid subscription required. Published overage unit rate; actual billing occurs in $9 per 100,000-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
Plus paid subscription required. Published overage unit rate; actual billing occurs in $100 per 1.25M-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
API terms allow use of content to provide and improve services and explicitly authorize training/improvement using consented cloning recordings, with opt-out and deletion requests available. Custom voice models remain on provider infrastructure. The privacy policy retains voice generation records until membership withdrawal, with separate statutory retention for transactions and access logs. No default no-training or zero-retention guarantee is established; exact inference region is not published in the reviewed material.
POST https://api.typecast.ai/v1/text-to-speech/compose. X-API-KEY required. Ordered multi-voice/model segments become a single binary audio response. Use a voice compatible with the selected model.
Regions
Exact inference region not established. Korean operator location and U.S. analytics subprocessors do not establish model residency.
Prices retain their published currencies and billing units below.
Published billing units · original currency · observed 16 Oct 2026
Category
Usage
Rate
Unit
API synthesis credits
Synthesized text1
1 credit
character
Lite overage
Additional synthesized text2
$0.09
1,000 characters
Plus overage
Additional synthesized text3
$0.08
1,000 characters
Rate conditions
One credit per text character. Monthly plan allowance and overage blocks apply. Compose bills combined text and no pause credits; a failed composition consumes no credits. Smart-context field billing details beyond the synthesized text meter are not independently verified.
Lite paid subscription required. Published overage unit rate; actual billing occurs in $9 per 100,000-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
Plus paid subscription required. Published overage unit rate; actual billing occurs in $100 per 1.25M-credit block, not individual 1,000-character charges. Included monthly credits and Studio plans are separate. Never convert this character rate into a model token price.
API terms allow use of content to provide and improve services and explicitly authorize training/improvement using consented cloning recordings, with opt-out and deletion requests available. Custom voice models remain on provider infrastructure. The privacy policy retains voice generation records until membership withdrawal, with separate statutory retention for transactions and access logs. No default no-training or zero-retention guarantee is established; exact inference region is not published in the reviewed material.
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.
Typecast API
ssfm-v21
Not verified
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How it works
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Minimum reusable input
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Lifetime & refresh
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How to enable it
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Usage fields
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Billing rules
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Limits & data boundaries
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 Oct 2026
Typecast API
ssfm-v21
Not verified
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How it works
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Minimum reusable input
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Lifetime & refresh
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How to enable it
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Usage fields
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Billing rules
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Limits & data boundaries
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 Oct 2026
Typecast API
ssfm-v21
Not verified
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How it works
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Minimum reusable input
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Lifetime & refresh
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How to enable it
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Usage fields
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Billing rules
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Limits & data boundaries
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 Oct 2026
Typecast API
ssfm-v21
Not verified
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How it works
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Minimum reusable input
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Lifetime & refresh
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
How to enable it
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Usage fields
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Billing rules
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
Limits & data boundaries
Not collected
No native inference cache contract established. Custom voice profiles, downloaded audio and retained generation history are not prompt caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 16 Oct 2026
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
Neosapience explicitly describes its history of proprietary speech model development. Current API selectors, licensed AWS software and historical generation labels are distinguished. Public TA2.0 avatar weights are a different model and are excluded.
