# Model Selectionw
We recommend starting with what you are trying to do, then selecting the best model from there.
### Image
Upscale an Image
Three tiers, each with a different tradeoff:
| | Gigapixel | Wonder | Bloom |
| --------------- | ---------------------------------------------------- | -------------------------------------------------------------------------- | ----------------------------------------------------- |
| **Tier** | Precision | Generative | Creative |
| **Description** | Precision Image Upscale & Enhancement | Generative Image Upscale & Enhancement | Creative Image Upscale & Enhancement |
| **Intent** | Improve resolution, preserve source characteristics. | Improve and restore by adding detail and texture, while preserving intent. | Transform with new, creative detail or stylization. |
| **Use Case** | General purpose upscaling & enhancement. | Upcaling, restoring and improving low resolution or compressed sources. | Upscaling, improving, and transforming GenAI sources. |
#### Gigapixel
| Model | Use Case |
| ----------------------------------------------------------------------------------- | ------------------------------------------------ |
| [`Standard 2`](/image-models/gigapixel/standard-2.md) | Default — works for most images |
| [`High Fidelity 2`](/image-models/gigapixel/high-fidelity-2.md) | Maximum source preservation |
| [`Low Resolution 2`](/image-models/gigapixel/low-resolution-2.md) | Very small inputs (thumbnails, web scrapes) |
| [`Art & CGI`](/image-models/gigapixel/art-and-cgi.md) | Illustrations, renders, game assets, digital art |
| [`Text & Shapes`](/image-models/gigapixel/text-and-shapes.md) | Documents, screenshots, text-heavy images |
| [`Recover Faces`](/image-models/gigapixel/recover-faces.md) | Portraits where facial detail is the priority |
| [`Transparent Image Upscale`](/image-models/gigapixel/transparent-image-upscale.md) | PNGs with alpha channels |
#### Wonder
| Model | Use Case |
| --------------------------------------------------------- | ------------------------------------------------- |
| [`Standard Max`](/image-models/wonder/standard-max.md) | Default generative upscaling |
| [`Wonder 2`](/image-models/wonder/wonder-2.md) | Stronger generation for more detail |
| [`Redefine: Realistic`](/image-models/wonder/redefine.md) | Generative enhancement grounded in photorealism |
| [`Redefine: Creative`](/image-models/wonder/redefine.md) | Generative enhancement with more creative freedom |
| [`Recover 3`](/image-models/wonder/recover-3.md) | Heavily degraded or damaged sources |
#### Bloom
| Model | Use case |
| --------------------------------------------------------- | -------------------------- |
| [`Bloom Creative`](/image-models/bloom/bloom-creative.md) | Maximum stylization |
| [`Bloom Realism`](/image-models/bloom/bloom-realism.md) | Creative but more grounded |
Sharpen or Deblur
**Standard** — use when you know the blur type:
| Model | Use case |
| --------------------------------------------------------------------------------- | ----------------------------------------------- |
| [`Sharpen`](/image-models/removal-and-cleanup/sharpen/standard.md) | General-purpose — good starting point |
| [`Sharpen Strong`](/image-models/removal-and-cleanup/sharpen/strong.md) | Noticeably soft images |
| [`Sharpen Natural`](/image-models/removal-and-cleanup/sharpen/natural.md) | Subtle sharpening that avoids over-processing |
| [`Sharpen Lens Blur 2`](/image-models/removal-and-cleanup/sharpen/lens-blur-2.md) | Optical blur from lens softness or missed focus |
| [`Sharpen Motion Blur`](/image-models/removal-and-cleanup/sharpen/motion-blur.md) | Blur from camera or subject movement |
| [`Sharpen Refocus`](/image-models/removal-and-cleanup/sharpen/refocus.md) | Slightly out-of-focus images |
| [`Portrait`](/image-models/removal-and-cleanup/sharpen/portrait.md) | Optimized for faces and skin texture |
| [`Wildlife`](/image-models/removal-and-cleanup/sharpen/wildlife.md) | Optimized for fur, feathers, scales |
**Generative** — use when the source is heavily degraded or you're unsure of the blur type:
| Model | Use case |
| ----------------------------------------------------------------------------- | -------------------------------------------------------------- |
| [`Super Focus 2`](/image-models/removal-and-cleanup/sharpen/super-focus-2.md) | Generative sharpening that recovers detail beyond the original |
Reduce Noise or Grain
Pick based on noise severity:
| Model | Use case |
| ----------------------------------------------------------------------------- | ----------------------------------------------------- |
| [`Denoise Normal`](/image-models/denoise/normal.md) | Light noise — well-lit conditions |
| [`Denoise Strong`](/image-models/denoise/strong.md) | Moderate noise — high ISO, low light |
| [`Denoise Extreme`](/image-models/denoise/extreme.md) | Heavy noise — very high ISO, underexposed |
| [`Dust & Scratch 2`](/image-models/removal-and-cleanup/dust-and-scratch-2.md) | Film scans or archival images with physical artifacts |
Remove Objects or Backgrounds
| Model | Use case |
| ----------------------------------------------------------------------------------- | ------------------------------------------------- |
| [`Background Removal`](/image-models/removal-and-cleanup/background-removal.md) | Remove the background entirely |
| [`Image Object Matting`](/image-models/removal-and-cleanup/image-object-matting.md) | Extract subject with transparency for compositing |
Correct Color or Lighting
| Model | Use case |
| ------------------------------------------------------------------------------ | --------------------------------------- |
| [`Balance Color`](/image-models/color-and-lighting/balance-color.md) | Auto white balance and color correction |
| [`Adjust Lighting 2`](/image-models/color-and-lighting/adjust-lighting-2.md) | Exposure and contrast adjustment |
| [`Image Colorization`](/image-models/color-and-lighting/image-colorization.md) | Convert black-and-white to color |
### Video
Upscale or Enhance a Video
Three families, each with a different tradeoff:
| | Proteus | Starlight | Astra |
| --------------- | ---------------------------------------------------- | -------------------------------------------------------------------------- | ----------------------------------------------------- |
| **Tier** | Precision | Generative | Creative |
| **Description** | Precision Video Upscale & Enhancement | Generative Video Upscale & Enhancement | Creative Video Upscale & Enhancement |
| **Intent** | Improve resolution, preserve source characteristics. | Improve and restore by adding detail and texture, while preserving intent. | Transform with new, creative detail or stylization. |
| **Use Case** | General purpose upscaling & enhancement. | Upcaling, restoring and improving archival and GenAI | Upscaling, improving, and transforming GenAI sources. |
#### Proteus
**Don't know where to start?** Use **Proteus**. Then narrow down if your source has specific characteristics:
| Model | Use case |
| ----------------------------------------------------------------------- | -------------------------------------------------------------------- |
| [`Proteus`](/video-models/proteus/proteus.md) | Default — handles most content well |
| [`Proteus Natural`](/video-models/proteus/proteus-natural.md) | Lighter processing, stays closest to the original |
| [`Artemis HQ / MQ / LQ`](/video-models/proteus/artemis.md) | Match to your source quality — HQ for clean footage, LQ for degraded |
| [`Artemis Aliasing/Moire`](/video-models/proteus/artemis.md) | Visible aliasing or moire patterns |
| [`Artemis Medium Halo / Strong Halo`](/video-models/proteus/artemis.md) | Halo artifacts from older upscaling or edge enhancement |
| [`Dione TV / DV`](/video-models/proteus/dione.md) | Interlaced broadcast (TV) or camcorder (DV) footage |
| [`Dione Robust`](/video-models/proteus/dione.md) | Interlaced content with heavy degradation |
| [`Dione Dehalo / Robust Dehalo`](/video-models/proteus/dione.md) | Interlaced content with halo artifacts |
| [`Gaia 2 (Animation)`](/video-models/proteus/gaia.md) | Animation and single-frame animation |
| [`Gaia HQ / CG`](/video-models/proteus/gaia.md) | High-quality or general CG and rendered content |
| [`Iris MQ / LQ`](/video-models/proteus/iris.md) | Legacy content at medium or low quality |
| [`Rhea`](/video-models/proteus/rhea.md) | General enhancement with natural texture |
| [`Theia Fine Tune Detail`](/video-models/proteus/theia.md) | Push detail enhancement — more aggressive |
| [`Theia Fine Tune Fidelity`](/video-models/proteus/theia.md) | Preserve original look — lighter touch |
#### Starlight
Quality and Fast subtiers. Use Quality when output matters most, Fast when speed or cost is the priority.
| Model | Subtier | Use case |
| ---------------------------------------------------------------------------------- | ------- | ------------------------------------------ |
| [`Starlight HQ`](/video-models/starlight/starlight-hq.md) | Quality | Maximum quality generative output |
| [`Starlight Precise 2`](/video-models/starlight/starlight-precise-2-deprecated.md) | Quality | Default for most generative work |
| [`Starlight Precise 1`](/video-models/starlight/starlight-precise-1-deprecated.md) | Quality | Highest fidelity to source |
| [`Starlight Sharp`](/video-models/starlight/starlight-sharp.md) | Quality | Emphasis on sharpness and edge detail |
| [`Starlight Fast 2`](/video-models/starlight/starlight-fast-2.md) | Fast | Best fast model — start here for speed |
| [`Starlight Fast 1`](/video-models/starlight/starlight-fast-1-deprecated.md) | Fast | Speed-optimized generative enhancement |
| [`Starlight Mini`](/video-models/starlight/starlight-mini.md) | Fast | Lightest — previews or cost-sensitive jobs |
#### Astra
| Model | Use case |
| ------------------------------------------------------ | ----------------------------- |
| [`Astra 1`](/video-models/astra/astra-1-deprecated.md) | Creative video transformation |
Reduce Noise or Grain
#### Denoise
Quality and Fast subtiers via the Nyx family:
| Model | Subtier | Use case |
| ----------------------------------------------------------------- | ------- | ---------------------------------------------- |
| [`Nyx`](/video-models/denoise/nyx.md) | Quality | Default — balanced noise reduction and detail |
| [`Nyx XL`](/video-models/denoise/nyx-xl.md) | Quality | Stronger reduction for heavy noise |
| [`Nyx High Fidelity`](/video-models/denoise/nyx-high-fidelity.md) | Quality | Maximum detail preservation, lighter reduction |
| [`Nyx Fast`](/video-models/denoise/nyx-fast.md) | Fast | Speed-optimized denoising |
Change Frame Rate or Create Slow-Mo
#### Frame Interpolation
Quality and Fast subtiers. Generates new frames to increase frame rate or create slow motion.
| Model | Subtier | Use case |
| --------------------------------------------------------- | ------- | ------------------------------------------------------- |
| [`Apollo`](/video-models/frame-interpolation/apollo.md) | Quality | Default — best all-around quality |
| [`Chronos`](/video-models/frame-interpolation/chronos.md) | Quality | Alternative algorithm for fast action or complex scenes |
| [`Aion`](/video-models/frame-interpolation/aion.md) | Quality | Additional option for specific content types |
| [`Apollo Fast`](broken://pages/dNHtFXtpCMkrraa22s9w) | Fast | Speed-optimized |
| [`Chronos Fast`](broken://pages/5rUaTuF8aMBBuSfjWfg0) | Fast | Speed-optimized alternative algorithm |
Deblur or Colorize
#### Video Utilities
| Model | Use case |
| --------------------------------------------------------------------------- | ----------------------------------------------------- |
| [`Themis 2`](/video-models/video-utilities/themis-2-motion-deblur.md) | Remove motion blur from camera shake or fast movement |
| [`Video Colorization`](/video-models/video-utilities/video-colorization.md) | Convert B\&W or desaturated footage to color |
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://developer.topazlabs.com/getting-started/model-selectionw.md?ask=
```
The question should be specific, self-contained, and written in natural language.
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.