Searchers use broad phrases such as AI video generation, AI image generation, “Veo 3,” “Veo 3.1,” “Grok video,” “GPT-2 image,” “Nano Banana,” “Seedream,” and “Wan.” Those phrases do not all describe models that are enabled in the same product. This guide separates the names that are currently visible in MotionVeo from names that are absent, ambiguous, or not applicable to the current generation catalog.
Current MotionVeo catalog at a glance
The table below is a capability map, not a league table. “Listed” means the user-facing label appeared in the dated connected catalog snapshot. “Not listed” means this page makes no access claim for that name.
| Model or search name | Type | Current status | Listed settings or clarification |
|---|---|---|---|
| Veo 3.1 Fast | Video | Listed | 1080p, 8 seconds, 16:9 or 9:16, source image, first frame, and up to 7 references. |
| Seedance 2.0 | Video | Listed | 1080p, 8 seconds, 16:9 or 9:16, source image, first frame, and up to 7 references. |
| Wan 2.6 | Video | Listed | 1080p, 8 seconds, 16:9 or 9:16, source image, first frame, and up to 7 references. |
| Seedream 4.0 | Image | Listed | 1080p, 2K, and 4K; 1:1, 16:9, 9:16, 4:3, and 3:4; source image listed. |
| Nano Banana 2 | Image | Listed | 1080p; 1:1, 16:9, 9:16, 4:3, and 3:4; source image listed. |
| Qwen Image | Image | Listed | 720p and 1080p; 1:1, 16:9, 9:16, 4:3, and 3:4; source image listed. |
| Veo 3 | Video search phrase | Not listed separately | The snapshot lists Veo 3.1 Fast, not a separate enabled model labeled Veo 3. |
| Grok | Search name | Not listed | No Grok generation model appears in this MotionVeo snapshot. This page does not claim access or integration. |
| GPT-2 | Search name | Not listed | GPT-2 does not appear as a MotionVeo image or video generation option in this snapshot. |
Video generation models: Veo 3.1, Seedance, and Wan
For short-form AI video generation, the current directory exposes Veo 3.1 Fast, Seedance 2.0, and Wan 2.6. The listed fields are currently the same: 1080p, an 8-second duration, 16:9 or 9:16 framing, source-image and first-frame support, and up to seven reference images.
Shared settings do not establish that the models produce identical results or that one is universally better. A video model can behave differently with the same prompt depending on subject complexity, camera movement, source image, references, and the destination format. Use the video model comparison for a focused capability table, then test the exact shot you need.
Video selection rule
- Need a short landscape or portrait scene? Start with any of the three listed video labels that your account can access.
- Need image continuity? Use the source-image, first-frame, or reference workflow that the live selector exposes.
- Need a quality verdict? Run the same prompt, ratio, and input through a small controlled test. Catalog settings alone cannot provide a ranking.
Image generation models: Seedream, Nano Banana, and Qwen
For AI image generation, MotionVeo currently lists Seedream 4.0, Nano Banana 2, and Qwen Image. Seedream 4.0 has the broadest listed resolution range at 1080p, 2K, and 4K. Nano Banana 2 lists 1080p. Qwen Image lists 720p and 1080p. All three currently list common landscape, portrait, square, and 4:3 aspect ratios plus source-image input.
“Nano Banana” is a common search shorthand, while the current MotionVeo display label is Nano Banana 2. The page uses the visible catalog label so readers can match the article to the Studio selector. It does not infer a provider relationship, benchmark score, or guaranteed image quality from the name.
An image model can also be the first step in an image-to-video workflow. Build a clean still, choose the final ratio, and then confirm that the selected video model accepts the source image. The image-to-video guide covers that handoff in more detail.
What about Veo 3, Grok, GPT-2, and other names?
Search demand is not the same as product availability. The current catalog does not show a separate enabled model labeled Veo 3, and it does not show Grok or GPT-2 as MotionVeo image or video generation options. This article mentions those names to answer real comparison questions, not to imply that they can be selected, purchased, or called through MotionVeo.
The same rule applies to any other model not shown in the live model directory. MotionVeo is an independent interface connected to a changing catalog. It does not claim ownership, endorsement, direct provider affiliation, or access to every model discussed across the wider AI market. Check the selector and the applicable provider terms before relying on a model in a production plan.
How to choose and test a model responsibly
A useful model comparison starts with the deliverable, not the headline. Define the output type, aspect ratio, source material, duration, and one visible creative goal. Then use the smallest fair test that can answer the question.
- Choose video or image first. Do not compare a still-image model with a short-video model as if they were interchangeable.
- Lock the format. Keep 16:9 or 9:16 fixed for a video comparison, or keep the same image ratio and resolution target for a still comparison.
- Keep the brief stable. Use the same prompt, source image, and reference set when comparing listed models.
- Inspect the complete result. For video, review motion, continuity, framing, faces, hands, text, and edges. For images, inspect composition, details, text, and the intended display size.
- Record one next change. Change one variable at a time so the result teaches you something about the workflow.
For prompt structure, see the AI video prompt guide. For a broader starting point, visit the AI video generator hub or the AI image generator hub.
Frequently asked questions
Which AI video models does MotionVeo currently list?
The catalog snapshot checked September 11, 2026 lists Veo 3.1 Fast, Seedance 2.0, and Wan 2.6 for video. Each currently lists 1080p, 8-second output, 16:9 and 9:16 framing, source-image and first-frame support, and up to 7 reference images.
Is Veo 3 available separately from Veo 3.1 in MotionVeo?
No separate enabled model labeled Veo 3 appeared in the snapshot. MotionVeo lists the user-facing label Veo 3.1 Fast. Verify the live Studio selector before a paid render because model names and availability can change.
Does MotionVeo offer Grok or GPT-2 for video or image generation?
Grok and GPT-2 are not listed in the current MotionVeo generation catalog. This page does not imply access, integration, provider affiliation, or future availability for an absent model.
Which AI image models are currently listed?
The snapshot lists Seedream 4.0, Nano Banana 2, and Qwen Image. Seedream 4.0 lists 1080p, 2K, and 4K; Nano Banana 2 lists 1080p; and Qwen Image lists 720p and 1080p.
Does this page rank the models from best to worst?
No. The table reports catalog fields and practical starting points, not a controlled quality benchmark. The best choice depends on the shot, input image, format, prompt, account availability, and your review of the result.
How can I verify a model before generating?
Open the MotionVeo Studio model selector immediately before rendering. Confirm the exact display label, output settings, account eligibility, and current price or quota before submitting a paid generation.