MuoGen · Choose for your project
Compare AI models for your images and videos
Find models compatible with your output and references. This comparison brings the MuoGen catalog settings into two tables you can read without an account.
Choose by requirement · Compare image models · Compare video models
Which models fit your planned output?
The following lists are derived from the capabilities in the tables. Use them to check a specific combination of settings before preparing your brief.
- A square image at 4K
- Nano Banana Pro, Nano Banana 2.1, Nano Banana 2
These models offer 1:1 and 4K together. Check the combination: a 4K option on its own does not establish square output support.
- A selectable transparent background
- GPT Image 2.5 Sunburst, GPT Image 2.5 Flare, GPT Image 2, GPT Image 1.5
These entries expose the transparent background setting. Also choose an output format suitable for your composition and inspect the resulting file.
- Multiple images in one request
- GPT Image 2.5 Sunburst, GPT Image 2.5 Flare, GPT Image 2, Grok Imagine Image 2.0, GPT Image 1.5, Qwen Image 3.0 Pro, Qwen Image 3.0, Grok Imagine Image Quality
The maximum count varies by model. This describes requested outputs; it is separate from the number of uploaded references.
- A new 20-second clip
- Seedance 2.5, Wan 3.0
Their output duration range includes 20 seconds. Allowed video reference duration remains a separate constraint.
- A 15-second video reference and a 30-second output
- Seedance 2.5
This combination must meet both reference and output limits, plus any combined limit stated in the table. With a 30-second combined limit, 15 seconds of video references leave at most 15 seconds for output.
- A new video guided by a reference clip
- Seedance 2.5, Seedance 2.0, Seedance 2.0 Fast, Seedance 2.0 Mini, MiniMax H3, Wan 3.0
These creation workflows accept video references. Check file counts, sizes, and individual and total reference durations.
- Edit an existing video
- Seedance 2.5 Edit
The Edit workflow uses a source clip. Duration and aspect ratio follow that clip rather than the controls for creating a new shot.
Compare image models
13 image models available in the catalog. Open “Aspect ratios by resolution” to check the combination you need and “Upload limits” to prepare references.
| Model | Inputs and references | Resolutions | Images per request | Settings and formats |
|---|---|---|---|---|
GPT Image 2.5 Sunburstgpt-image-2.5-sunburstView settings | Text prompt + images 16 reference images maximum Upload limits
| 1K, 2K, 4K | 1–10 images | Quality levels: low, medium, high, xhigh, max Selectable formats: png, jpeg, webp Background modes: transparent, opaque, auto Prompt: 32000 characters maximum Initial settings: 1K, 1:1, high, png, auto Aspect ratios by resolution
|
GPT Image 2.5 Flaregpt-image-2.5-flareView settings | Text prompt + images 16 reference images maximum Upload limits
| 1K, 2K, 4K | 1–10 images | Quality levels: low, medium, high, xhigh, max Selectable formats: png, jpeg, webp Background modes: transparent, opaque, auto Prompt: 32000 characters maximum Initial settings: 1K, 1:1, high, png, auto Aspect ratios by resolution
|
GPT Image 2gpt-image-2View settings | Text prompt + images 16 reference images maximum Upload limits
| 1K, 2K, 4K | 1–10 images | Quality levels: low, medium, high Selectable formats: png, jpeg, webp Background modes: transparent, opaque, auto Prompt: 32000 characters maximum Initial settings: 1K, 1:1, high, png, auto Aspect ratios by resolution
|
Nano Banana Progemini-3-pro-imageView settings | Text prompt + images 14 reference images maximum Upload limits
| 1K, 2K, 4K | 1 image | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Grok Imagine Image 2.0grok-imagine-image-2.0View settings | Text prompt + images 3 reference images maximum Upload limits
| 1K, 2K | 1–10 images | Quality levels: low, medium Prompt: 30000 characters maximum Initial settings: 1K, 1:1, low Aspect ratios by resolution
|
Nano Banana 2.1gemini-nano-banana-2.1View settings | Text prompt + images 14 reference images maximum Upload limits
| 1K, 2K, 4K | 1 image | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Nano Banana 2gemini-3.1-flash-imageView settings | Text prompt + images 14 reference images maximum Upload limits
| 1K, 2K, 4K | 1 image | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
GPT Image 1.5gpt-image-1.5View settings | Text prompt + images 16 reference images maximum Upload limits
| 1K | 1–10 images | Quality levels: low, medium, high Selectable formats: png, jpeg, webp Background modes: transparent, opaque, auto Prompt: 32000 characters maximum Initial settings: 1K, 1:1, low, png, auto Aspect ratios by resolution
|
Nano Banana 2 Litegemini-3.1-flash-lite-imageView settings | Text prompt + images 14 reference images maximum Upload limits
| 1K | 1 image | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Qwen Image 3.0 Proqwen-image-3.0-proView settings | Text prompt + images 3 reference images maximum Upload limits
| 1K, 2K | 1–6 images | Additional controls: negative prompt, seed, prompt extension, thinking Prompt: 2000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Qwen Image 3.0qwen-image-3.0View settings | Text prompt + images 3 reference images maximum Upload limits
| 1K, 2K | 1–6 images | Additional controls: negative prompt, seed, prompt extension, thinking Prompt: 2000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Grok Imagine Image Qualitygrok-imagine-image-qualityView settings | Text prompt + images 3 reference images maximum Upload limits
| 1K, 2K | 1–10 images | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Nano Bananagemini-2.5-flash-imageView settings | Text prompt + images 14 reference images maximum Upload limits
| 1K | 1 image | No selectable quality, format or background mode Prompt: 30000 characters maximum Initial settings: 1K, 1:1 Aspect ratios by resolution
|
Source catalog checked on: .
Compare video models
9 video models available in the catalog. Output and reference durations are shown separately. The Edit workflow follows its source clip.
| Model | Inputs and references | Resolutions | Output duration | Settings and formats |
|---|---|---|---|---|
Seedance 2.5seedance-2.5View settings | Text prompt image: 30 video: 10 audio: 10 Maximum reference counts Upload limits
| 1080p, 720p, 480p | 4–30 s Initial duration: 10 s | Initial resolution: 720p Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive Audio on/off control exposed Prompt: 30000 characters maximum Video extension workflow Audio reference without an image or video accepted |
Seedance 2.5 Editseedance-2.5-editView settings | Text prompt image: 30 video: 10 audio: 10 Maximum reference counts Upload limits
| 1080p, 720p, 480p | Follows the source video (Edit) | Initial resolution: 720p Source video aspect ratio Audio on/off control exposed Prompt: 30000 characters maximum |
Seedance 2.0seedance-2.0View settings | Text prompt image: 9 video: 3 audio: 3 Maximum reference counts Upload limits
| 4k, 1080p, 720p, 480p | 4–15 s Initial duration: 5 s | Initial resolution: 720p Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive Audio on/off control exposed Prompt: 30000 characters maximum |
Seedance 2.0 Fastseedance-2.0-fastView settings | Text prompt image: 9 video: 3 audio: 3 Maximum reference counts Upload limits
| 720p, 480p | 4–15 s Initial duration: 5 s | Initial resolution: 720p Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive Audio on/off control exposed Prompt: 30000 characters maximum |
Seedance 2.0 Miniseedance-2.0-miniView settings | Text prompt image: 9 video: 3 audio: 3 Maximum reference counts Upload limits
| 720p, 480p | 4–15 s Initial duration: 5 s | Initial resolution: 720p Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive Audio on/off control exposed Prompt: 30000 characters maximum |
MiniMax H3minimax-h3View settings | Text prompt image: 30 video: 10 audio: 10 Maximum reference counts Upload limits
| 2K, 768P | 4–15 s Initial duration: 5 s | Initial resolution: 768P Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive No separate audio on/off control Prompt: 7000 characters maximum |
MiniMax H3 Maxminimax-h3-maxView settings | Text prompt image: 2 Maximum reference counts Upload limits
| 768P, 480P | 5–15 s Initial duration: 5 s | Initial resolution: 768P Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive No separate audio on/off control Prompt: 7000 characters maximum First and last frame images |
Grok Imagine 1.5grok-imagine-1.5View settings | Text prompt image: 14 Maximum reference counts Upload limits
| 1080p, 720p, 480p | 1–15 s Initial duration: 8 s | Initial resolution: 720p Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3 Audio on/off control exposed Prompt: 30000 characters maximum |
Wan 3.0wan-3.0View settings | Text prompt image: 10 video: 5 audio: 5 Maximum reference counts Upload limits
| 1080P, 720P, 480P | 2–30 s Initial duration: 5 s Video references + output ≤ 30 s With 15 s of video references, output is limited to 15 s. Audio references remain counted separately. | Initial resolution: 1080P Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, adaptive Audio on/off control exposed Prompt: 20000 characters maximum Audio reference without an image or video accepted |
Source catalog checked on: .
How should I use these differences?
- Define the deliverable: image or video, aspect ratio, resolution and duration. Check the settings together before preparing references.
- Count references and check their files. A total limit can be reached before the maximum file count; audio and video reference durations have separate totals.
- Distinguish exposed controls from outcomes to review. A variant name, resolution or reference count does not measure output quality, processing time or reference fidelity.
- Read the model guide and actual settings. Prepare a focused brief, then review the quote before authorizing an available generation.
The tables use the same public catalogs as the MuoGen workspaces. They include every model marked available and the controls exposed by those catalogs.
On mobile, scroll each table horizontally. Table regions can receive keyboard focus; open the limits and ratio lists using their headings.
What is the current generation availability?
The catalog and this comparison are open to browse. Paid generation in the public browser workspaces is not yet open. When enabled, it will require a verified account and paid credits. Reading these tables or opening settings does not start a generation.
Prepare your project with MuoGen
Updated: