Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Pika / Pika Audio

Pika Speech

Text-to-speech with reference-voice cloning and 76 published voice presets.

availableClosed weightsText-to-speech
Official documentation (opens in a new tab)Reviewed 13 Oct 2026
11 of 22 specification fields collected
Text input limit15,000 charactersSee specification notes below
LanguagesNot collectedAwaiting source review
Input billingSee ratesPika API Club
Output billingSee ratesPublished units and conditions below

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
Reasoning
Not collected
Tool calling
Not collected
Structured output
Not collected
Streaming
Not collected

The technical article discusses a streaming variant, but the current API Club media route is asynchronous. No accessible streaming speech endpoint has been mapped to that variant.

Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2026-08-14Source for Release date (opens in a new tab)

Family announcement explicitly states all four models are available. Individual technical articles followed on August 18.

Parameters
3 billion parameters (research architecture)Source for Parameters (opens in a new tab)
Architecture
48-block flow-matching diffusion transformer; the API identifies an eight-step DMD generatorSource for Architecture (opens in a new tab)

The API implementation description is separately captured in its request contract.

Languages
Not collected
Context caching
Not collected
Maximum prompt characters
15,000 charactersSource for Maximum prompt characters (opens in a new tab)
Output duration
Up to 5 minutes per requestSource for Output duration (opens in a new tab)
Voice cloning
SupportedSource for Voice cloning (opens in a new tab)

Reference audio requires explicit voice_consent_attested. Voice conditioning is not weight fine-tuning.

Voice controls
76 presets or a reference voice; seed controlSource for Voice controls (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.

API model ID
pika/pika-audio/pika-speech
Access
API Club account, $10/month membership, usage balance and API key required. Public contract reviewed; no paid generation request was made.
Regions
Not established; public availability does not establish processing location or data residency.
API endpoint URL
https://api.dev.pika.art/v1/media/pika/pika-audio/pika-speechSource for API endpoint URL (opens in a new tab)
Request mode
Asynchronous media jobs with status polling and optional webhooksSource for Request mode (opens in a new tab)
Streaming
Not supportedSource for Streaming (opens in a new tab)

Applies to the media surface, not PikaStream or hosted third-party chat APIs.

Billing conditions
Published Pika API Club usage rate, including its stated 5% platform fee; do not add that fee again. Separate $10/month membership required. No web-plan, subscription-credit, fal or token-price conversion. Most-specific resolution/duration tier overrides a less-specific default.Source for Billing conditions (opens in a new tab)
Seed control
Integer seed; default -1Source for Seed control (opens in a new tab)
Maximum prompt characters
15,000 charactersSource for Maximum prompt characters (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 13 Oct 2026
CategoryUsageRateUnit
AudioPika Speech · Standard1$0.01minute

Rate conditions

  1. Published Pika API Club usage rate, including its stated 5% platform fee; do not add that fee again. Separate $10/month membership required. No web-plan, subscription-credit, fal or token-price conversion. Most-specific resolution/duration tier overrides a less-specific default. Published qualifiers: Standard.
Pricing source (opens in a new tab)

Data & enterprise terms

Pika developer terms apply to direct access. fal and other integration terms may additionally apply. Retention, training opt-out and regional enterprise guarantees have not been established for this endpoint.

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.

Pika API Club

pika/pika-audio/pika-speech
Not verified
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.

Voice choices

76 voices

Choose a voice within Pika Speech. These are API voice identifiers, not additional model releases. Source links document each voice.

Published voice catalog
Voice / API IDLanguage & accentDelivery
Deep Trailer Bassdeep_trailer_bassNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Warm Honey Drawlwarm_honey_drawlNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Fast Talking Wiseguyfast_talking_wiseguyNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Graceful Elder Storytellergraceful_elder_storytellerNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Chill Surfer Dudechill_surfer_dudeNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Bubbly It Girlbubbly_it_girlNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Roaring Locker Room Coachroaring_locker_room_coachNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.Voice source (opens in a new tab)
Calm Documentary Narratorcalm_documentary_narratorNot specifiedPublished preset ID; display name formatted from that ID. Per-preset language and accent are not specified in the contract.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

Official Pika release and API documents identify the native model or explicitly labeled managed service. Pika reserves its service intellectual property and prohibits reverse engineering and model extraction. Closed-weight classification is based on this proprietary hosted distribution and the absence of a published weight license for the named release, not merely on paid access. Hosted third-party models are excluded.

Classification evidence (opens in a new tab)

Lifecycle dates

Audio family available in API Club14 Aug 2026Source (opens in a new tab)
Individual technical article18 Aug 2026Source (opens in a new tab)

Sources reviewed