> For the complete documentation index, see [llms.txt](https://developer.topazlabs.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.topazlabs.com/video-models/upscaling-and-enhancement.md).

# Upscaling & Enhancement

Precision, Generative, and Creative Upscaling for Video

### Best Models

<table data-view="cards"><thead><tr><th></th><th></th><th></th><th data-hidden data-card-cover data-type="image">Cover image</th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><a class="button secondary medium">Precision</a></td><td><strong>Proteus</strong></td><td>The default choice for video enhancement.</td><td><a href="https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FUrxNNr6PLLPjh6Z9XyhU%2FProteus_(Ex_1)_slider.jpg?alt=media&#x26;token=ee198405-3d9f-458e-b946-a8272daf07b0">https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FUrxNNr6PLLPjh6Z9XyhU%2FProteus_(Ex_1)_slider.jpg?alt=media&#x26;token=ee198405-3d9f-458e-b946-a8272daf07b0</a></td><td><a href="https://developer.topazlabs.com/video-models/upscaling-and-enhancement/precision/proteus.md">Proteus</a></td></tr><tr><td><a class="button secondary medium">Generative</a></td><td><strong>Starlight Precise 2.6</strong></td><td>Our highest-quality generative video upscaler.</td><td><a href="https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FTzrLBGS7de5x7Zd0EGiN%2Fslp-4k-upscale-example_slider.jpg?alt=media&#x26;token=fd897da0-97a2-4cee-9f9c-bbf1e1da2428">https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FTzrLBGS7de5x7Zd0EGiN%2Fslp-4k-upscale-example_slider.jpg?alt=media&#x26;token=fd897da0-97a2-4cee-9f9c-bbf1e1da2428</a></td><td><a href="https://developer.topazlabs.com/video-models/upscaling-and-enhancement/generative/starlight-precise-2.6.md">Starlight Precise 2.6</a></td></tr><tr><td><a class="button secondary medium">Creative</a></td><td><strong>Astra 2</strong></td><td>Creative, prompt-guided upscaling for GenAI video.</td><td><a href="https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FYNF6oN4TP7jjpzsInYws%2Fvideo_12_slider.jpg?alt=media&#x26;token=10425eeb-84ba-4f43-bfa8-f2b4a9763863">https://4007072434-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FJkSwaYgAsTQke14gpXAF%2Fuploads%2FYNF6oN4TP7jjpzsInYws%2Fvideo_12_slider.jpg?alt=media&#x26;token=10425eeb-84ba-4f43-bfa8-f2b4a9763863</a></td><td><a href="https://developer.topazlabs.com/video-models/upscaling-and-enhancement/creative/astra-2.md">Astra 2</a></td></tr></tbody></table>

### Upscaling Tiers

These are our three tiers of upscaling, ordered from most precise to most generated detail. Here's more information on how to choose the best tier for your needs:

* [**Precision Upscaling**](https://developer.topazlabs.com/video-models/upscaling-and-enhancement/precision.md) **—** Faithful to the source. Upscales, denoises, and sharpens without generating new detail; best for camera footage, broadcast, and anything that must match the original. **Proteus** is our recommended general-purpose model for this tier.
* [**Generative Upscaling**](https://developer.topazlabs.com/video-models/upscaling-and-enhancement/generative.md) **—** Diffusion models that rebuild realistic faces, texture, and text frame by frame; best for GenAI, archival, and heavily degraded video. **Starlight Precise 2.6** is our recommended general-purpose model for this tier.
* [**Creative Upscaling**](https://developer.topazlabs.com/video-models/upscaling-and-enhancement/creative.md) **—** Reimagines the footage with bold new detail and optional prompts; best for GenAI and stylized video. **Astra 2** is our recommended general-purpose model for this tier.

### Pricing

Each tier has its own base rate, priced by the number of output frames at the output resolution:

`credits = ceil(output frames ÷ frames per credit)`

Credits per minute are estimates at 30 fps. Higher output frame rates use more frames and cost more.

|                                                                   | Frames per credit (1080p) | Frames per credit (4K) | Credits per minute, 1080p | Credits per minute, 4K |
| ----------------------------------------------------------------- | ------------------------- | ---------------------- | ------------------------- | ---------------------- |
| **Precision** (Proteus and the Precision models)                  | —                         | —                      | 8                         | 29                     |
| **Generative** (Starlight Precise 2.6, Fast 3, Mini, Precise 2.5) | 26.0417                   | 11.9179                | 70                        | 152                    |
| **Creative** (Astra 2)                                            | 10                        | 6                      | 180                       | 300                    |

A few models are priced off their tier's base rate: Gaia 2 is 300 frames per credit at 1080p (100 at 4K), and Starlight Fast 2 is 52.0833 frames per credit at 1080p (23.8357 at 4K). Each model page lists its exact rate and estimated costs for common durations and frame rates. For estimates on your own files, use the [Credit Calculator](https://developer.topazlabs.com/credit-calculator).

### Endpoints

| Endpoint                                  | Use when                                                                                                                                                                                                                                       |
| ----------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `https://api.topazlabs.com/video/`        | You can supply full source metadata and want a credit estimate before uploading. Follow with `PATCH /video/{requestId}/accept` and `PATCH /video/{requestId}/complete-upload/`.                                                                |
| `https://api.topazlabs.com/video/express` | You only know the container. Returns upload URLs right away, with no accept or complete-upload steps. **Note that this endpoint is only available for GAN models (i.e. most creative and generative upscaling models will not be supported).** |

Every model in this family works on both endpoints. Requests are JSON: put the model's code in a `filters` object, along with any of its model parameters. Poll `GET /video/{requestId}/status` and download the result from the URL in the status response.

```bash
curl -X POST "https://api.topazlabs.com/video/express" \
  -H "X-API-Key: $TOPAZ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "source": { "container": "mp4" },
    "filters": [ { "model": "slp-2.6" } ],
    "output": { "resolution": { "width": 3840, "height": 2160 } }
  }'
```

### Parameters

Request parameters (`source`, `output`, `destination`, `notifications`, and supported encoders and profiles) are the same for every video model and are listed on the Request Parameters page. Each model's filter parameters are listed on its model page.

### Limitations

Upscaling models are limited by the output encoder's maximum resolution; see Supported encoders and profiles on the Request Parameters page. Proteus Natural and Gaia 2 upscale 2× only, and Rhea upscales 4×. Dione models expect interlaced input.

### Helpful Resources

<table data-view="cards"><thead><tr><th></th><th></th><th></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><i class="fa-play">:play:</i></td><td><h4>Video Quickstart</h4></td><td>Quickly get started with our Video API.</td><td><a href="https://developer.topazlabs.com/getting-started/video-quickstart.md">Video Quickstart</a></td></tr><tr><td><i class="fa-file-brackets-curly">:file-brackets-curly:</i></td><td><h4>API Reference</h4></td><td>View our API reference to view the entire schema.</td><td><a href="https://developer.topazlabs.com/reference/">https://developer.topazlabs.com/reference/</a></td></tr><tr><td><i class="fa-square-root-variable">:square-root-variable:</i></td><td><h4>Model Pricing</h4></td><td>Learn how we price our models.</td><td><a href="https://developer.topazlabs.com/getting-started/model-pricing.md">Model Pricing</a></td></tr></tbody></table>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://developer.topazlabs.com/video-models/upscaling-and-enhancement.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
