Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Yandex / Yandex SpeechKit Brand Voice

SpeechKit Brand Voice Self Service

Custom voice offering documented in Yandex Cloud Kazakhstan, with customer-prepared training data and access by arrangement.

Speech & voice
restrictedClosed-source managed componentText-to-speech
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
10 of 24 specification fields collected
Text input limitAPI v3: 250 normally; 5,000 with unsafe/streaming mode, subject to voice access.See specification notes below
Languages2 listedFull language list in specifications
Input billingNot collectedYandex SpeechKit Brand Voice
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 input tokens
Not collected
Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected
Streaming
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Languages
Russian, KazakhSource for Languages (opens in a new tab)

Other languages require provider confirmation.

Context caching
Not collected
Input modalities
Text, Audio referenceSource for Input modalities (opens in a new tab)

Audio reference is used for pattern-based synthesis.

Output modalities
AudioSource for Output modalities (opens in a new tab)
Voice cloning
SupportedSource for Voice cloning (opens in a new tab)

Creates a customer-specific voice from training recordings; not a public stock voice.

Audio sample rates
Training audio: 48 kHz; synthesis: 22 kHz.Source for Audio sample rates (opens in a new tab)
Maximum prompt characters
API v3: 250 normally; 5,000 with unsafe/streaming mode, subject to voice access.Source for Maximum prompt characters (opens in a new tab)
Output duration
Normal API v3 requests: 24 seconds; long-form synthesis uses unsafe/streaming mode.Source for Output duration (opens in a new tab)
Endpoint availability
Provider access or voice provisioning required; not an unrestricted stock model.Source for Endpoint availability (opens in a new tab)
Voice references
Customer supplies recordings and matching transcripts. Mono WAV at 48 kHz/16-bit PCM; individual clips at most 15 seconds. Dataset size depends on the target voice and use case.Source for Voice references (opens in a new tab)
Deployment options
Provisioned voice_id in Yandex Cloud; SpeechKit Hybrid delivery may be arranged.Source for Deployment options (opens in a new tab)
Usage metering
Not collected

No USD rate specific to the regional Self Service offering verified. Do not reuse Lite or Premium prices.

Access & pricing

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

Yandex SpeechKit Brand Voice

Provider source (opens in a new tab)
API model ID
Not collected

No shared immutable model ID is published for this route; do not invent one.

Access
SpeechKit API v3 using a provisioned private voice ID; contact provider for entitlement.
Regions
Contract and deployment dependent; no universal processing region inferred.

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

Data & enterprise terms

Model-specific retention and training-use contract not collected.

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.

Yandex SpeechKit Brand Voice

Not verified

No exact-model prompt cache or cache discount verified. Repeat requests are not assumed free.

How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Not collected
Limits & data boundaries
Not collected

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

Caching reviewed 14 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

Hosted translation or custom voice deployment offering. The served component is not published as a downloadable checkpoint in the reviewed sources. This is not an enumeration of customer-private weights or a claim of one immutable foundation model.

Classification evidence (opens in a new tab)

Sources reviewed