Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Tavus / Sparrow

Sparrow-1

Audio-native recurrent timing model that predicts conversational floor ownership from streaming speech.

availableClosed weightsTurn detection
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
5 of 16 specification fields collected
Context windowNot collectedAwaiting source review
Maximum outputNot collectedAwaiting source review
Input billingNot collectedTavus CVI
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

Exact model-version information has not been established; unknown does not mean unsupported.

Maximum output
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Context caching
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Reasoning
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Tool calling
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Structured output
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Batch processing
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Fine-tuning
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Parameters
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Architecture
Audio-native recurrent conversational-floor predictor with persistent streaming stateSource for Architecture (opens in a new tab)
Knowledge cutoff
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Languages
Not collected

Exact model-version information has not been established; unknown does not mean unsupported.

Output modalities
Control signalsSource for Output modalities (opens in a new tab)

Turn ownership and timing decisions, not a transcript or synthesized speech.

Access & pricing

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

API model ID
sparrow-1
Access
Set layers.conversational_flow.turn_detection_model=sparrow-1 when creating or updating a PAL.
Regions
Privacy policy describes US processing and possible transfers to service providers in other countries. Exact model-serving location is not established.
frameInterval
40 ms per streaming state updateSource for frameInterval (opens in a new tab)

Input cadence, not full conversational response latency.

timingControls
turn_taking_patience and pal_interruptibility support low, medium and high.Source for timingControls (opens in a new tab)
Primary voice isolation
voice_isolation supports near (default) and off.Source for Primary voice isolation (opens in a new tab)

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

Data & enterprise terms

Default privacy terms permit service improvement and AI research/training. Biometric retention is limited by the stated purpose or one year after last interaction, whichever comes first. Enterprise zero-retention arrangements are separate; no universal zero-retention promise is inferred.

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.

Tavus CVI

sparrow-1
Not verified

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

How it works
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

Minimum reusable input
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

Lifetime & refresh
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

How to enable it
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

Usage fields
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

Billing rules
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

Limits & data boundaries
Not collected

Internal streaming state, saved faces and conversation memories do not establish billable prompt caching, cache lifetime or discounts.

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

Caching reviewed 14 Oct 2026

What the measurements say

Floor-transfer precision

Tavus · provider-reported
1fraction

Provider test of 28 challenging conversational samples. Correct detection allows a 400 ms pre-end grace window; latency measured after user stops speaking. Not a population-wide accuracy guarantee.

Evaluation source (opens in a new tab)

Floor-transfer recall

Tavus · provider-reported
1fraction

Provider test of 28 challenging conversational samples. Correct detection allows a 400 ms pre-end grace window; latency measured after user stops speaking. Not a population-wide accuracy guarantee.

Evaluation source (opens in a new tab)

Interruptions

Tavus · provider-reported
0count

Provider test of 28 challenging conversational samples. Correct detection allows a 400 ms pre-end grace window; latency measured after user stops speaking. Not a population-wide accuracy guarantee.

Evaluation source (opens in a new tab)

Median response latency

Tavus · provider-reported
55ms

Provider test of 28 challenging conversational samples. Correct detection allows a 400 ms pre-end grace window; latency measured after user stops speaking. Not a population-wide accuracy guarantee.

Evaluation source (opens in a new tab)

Mean response latency

Tavus · provider-reported
292ms

Provider test of 28 challenging conversational samples. Correct detection allows a 400 ms pre-end grace window; latency measured after user stops speaking. Not a population-wide accuracy guarantee.

Evaluation source (opens in a new tab)

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

Named model developed by Tavus in its proprietary hosted/research stack. Reviewed primary documentation provides no downloadable weights for this model; service terms restrict source derivation and independent avatar use. Open SDKs and third-party models in the surrounding pipeline are separate.

Classification evidence (opens in a new tab)

Lifecycle dates

Named release announced13 Jan 2026Source (opens in a new tab)

Sources reviewed