Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Topaz Labs / Topaz starlight

Starlight Fast 3

Third-generation Starlight Fast video enhancement; its documented frame tariff differs from Fast 2.

Video processing
availableClosed weightsVideo processing
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
4 of 18 specification fields collected
Output resolutionNot collectedAwaiting source review
Clip durationNot collectedAwaiting source review
Input billingSee ratesTopaz Labs API
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

No model-specific public value established in reviewed documentation.

Maximum output
Not collected

No model-specific public value established in reviewed documentation.

Context caching
Not collected

No model-specific public value established in reviewed documentation.

Reasoning
Not collected

No model-specific public value established in reviewed documentation.

Tool calling
Not collected

No model-specific public value established in reviewed documentation.

Structured output
Not collected

No model-specific public value established in reviewed documentation.

Streaming
Not collected

No model-specific public value established in reviewed documentation.

Batch processing
Not collected

No model-specific public value established in reviewed documentation.

Fine-tuning
Not collected

No model-specific public value established in reviewed documentation.

Parameters
Not collected

No model-specific public value established in reviewed documentation.

Architecture
Not collected

No model-specific public value established in reviewed documentation.

Knowledge cutoff
Not collected

No model-specific public value established in reviewed documentation.

Languages
Not collected

No model-specific public value established in reviewed documentation.

Release date
Not collected

No model-specific public value established in reviewed documentation.

Input modalities
VideoSource for Input modalities (opens in a new tab)
Output modalities
VideoSource for Output modalities (opens in a new tab)
Supported API routes
https://api.topazlabs.com/video/Source for Supported API routes (opens in a new tab)
Published price
26.0417 frames/credit at 1080p; 11.9179 frames/credit at 4KSource for Published price (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.

Topaz Labs API

Provider source (opens in a new tab)
API model ID
slf-3
Access
API-key authenticated https://api.topazlabs.com/video/. Processing consumes credits; no paid call was performed during review.
Regions
Processing region not established in reviewed model documentation.
Request size limit
500 MB request-body limitSource for Request size limit (opens in a new tab)

This is an API request limit. The multipart video upload workflow does not establish a universal 500 MB maximum source-video file.

Request quotas
Load-dependent; HTTP 429 responses require retry/backoff. No fixed account RPM limit established.Source for Request quotas (opens in a new tab)
Processing workflow
Create job metadata, accept, upload parts, complete upload, then poll processing status.Source for Processing workflow (opens in a new tab)
Model controls
inputFrameCount integer optional default="null", inputFrameRate decimal optional default="null", inputHeight integer optional default=1080, inputWidth integer optional default=1920, outputHeight integer optional default=1080, outputWidth integer optional default=1920, videoBitDepth integer optional default="null", videoCodec string optional default="null", videoProfile string optional default="null", watermark boolean optional default="false"Source for Model controls (opens in a new tab)

Exact documented field names and type/range lines; optional descriptions are omitted. These controls are not weight fine-tuning.

Training data policy
Not collected

The reviewed general privacy policy does not establish a model-specific prohibition on training use.

Data retention
Not collected

No precise uploaded-media retention duration established for this model.

Content provenance
Provider FAQ says API outputs are C2PA-compliant.Source for Content provenance (opens in a new tab)

Provider statement, not an independently verified credential check.

Live service catalog
Listed in public supportedModels snapshotSource for Live service catalog (opens in a new tab)

Observed 2026-10-14. A listing is not a paid inference test; absence does not prove retirement. Explicit model-page deprecations remain in place.

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 14 Oct 2026
CategoryUsageRateUnit
Topaz media APIFrame-based processing11 credit26.0417 output frames at 1080p
Topaz media APIFrame-based processing11 credit11.9179 output frames at 4K

Rate conditions

  1. Published model-specific credit pricing. No conversion to USD, token cost or cost per question is inferred. Account terms, rounding and processing options can affect the bill. Provider publishes frames per credit. The displayed decimal precision is retained; this is not a fixed video-second charge.
Pricing source (opens in a new tab)

Data & enterprise terms

The general privacy policy retains personal information for service, business and legal purposes. No model-specific training opt-out, fixed media-retention period or selectable processing region is established by this policy.

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.

Topaz Labs API

slf-3
Not verified

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

How it works
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

Minimum reusable input
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

Lifetime & refresh
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

How to enable it
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

Usage fields
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

Billing rules
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

Limits & data boundaries
Not collected

No documented prompt-cache contract or discounted cache tariff. Saved files, processing jobs and local model downloads do not establish prompt caching.

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

The provider identifies these as its internally developed AI models. Its published proprietary software terms retain intellectual-property rights and prohibit redistribution and reverse engineering. API selectors are recorded without claiming the weights are openly licensed.

Classification evidence (opens in a new tab)

Sources reviewed

Topaz faq (opens in a new tab)Reviewed 14 Oct 2026