Below you’ll find answers to some of the most common questions about uploading, generating, and managing your images and videos. These FAQs cover everything from supported file formats and recommended resolutions to storage, privacy, and content management, helping you get the best results and keep your creations secure.
Uploading reference images and video
What image formats can be uploaded?
You can upload in any format. We convert your file into the correct format in the background.
Can I upload images with transparent backgrounds?
Currently, images with transparent elements aren't supported by our AI Image-to-Image or Image-to-Video generators. If you've encountered this error when uploading an image that shouldn't have any transparency, please try saving the image as a .JPEG instead of a .PNG and uploading the .JPEG instead.
What is the optimal image resolution for uploading?
The minimum image resolution is 0.5 MP.
Recommended: Use 1 MP or higher (e.g. 1280×720 px).
What image file size can be uploaded?
The image file uploaded must be under 10 MB.
What aspect ratio do the images I upload need to have?
You can upload images with an aspect ratio between 0.4:1 and 2.5:1.
How many reference images can I upload when generating a video?
The ability to use a start frame or a start and end frame is dependent on the model.
How many images can I use as a reference when generating an image?
The number of reference images you can upload when generating an image is dependent on the model. This will be indicated in settings option in the prompt area. To find out more about a specific model, find details about each model here.
What’s the best image size for transforming one image into another image?
There are no strict size requirements for image sizes, but for best results, upload an image that's at least 1 megapixel (e.g. 1280x720 px).
What file type, duration, size, and resolution does my reference video clip need to be for Kling 3.0 Motion Control and Kling 03 Video Edit?
Supported file types: MP4, MOV
Size: Total size under 200MB
Resolution: 720-2160px (720p-4K)
How many reference video clips can I upload with Seedance 2.0?
You can upload up to 3 reference video clips totaling 15 seconds.
How many reference video clips can I upload with Seedance 2.5?
You can upload up to 10 reference video clips totaling 30 seconds.
What file type, duration, size, and resolution do my reference video clips need to be for Seedance 2.0?
Supported file types: MP4, MOV
Durations: Up to 3 videos, combined duration must be between 2 and 15 seconds,
Size: Total size under 50 MB.
Resolution: Each video must be between ~480p (640x640) and ~720p (834x1112) in resolution.
What file type, duration, size, and resolution do my reference video clips need to be for Seedance 2.5?
Supported file types: MP4, MOV
Durations: Up to 10 videos, combined duration must be between 2 and 30 seconds
Size: Total size under 50 MB
Resolution: Each video must be between ~480p (640x640) and ~720p (834x1112) in resolution.
How many references can I upload for Seedance 2.0?
- Image Inputs -Up to 9 images
- Video Inputs - 3 clips, 15 seconds in total
- Audio Inputs -3 files, 15 seconds in total
You can upload up to 15 input files in total.
How many references can I upload for Seedance 2.5?
- Image Inputs — Up to 30 images
- Video Inputs — Up to 10 clips, 30 seconds in total
- Audio Inputs — Up to 10 files
You can upload up to 50 input files in total.
Is Seedance 2.5's Video Reference (Reference-to-Video) feature covered by Unlimited?
No — Unlimited covers Text-to-Video and Image-to-Video generations only. Using Video Reference (Reference-to-Video) with Seedance 2.5 requires credits regardless of your plan.
What happens if my uploaded video clip is longer than the model’s duration limit?
If your video (or the combined total of multiple video references) is longer than the model’s allowed duration, you’ll see a warning message in the prompt box and a Trim video link on the reference. Click it to open the video trimmer, drag the selection handles to choose the section you want to use, and confirm — no need to re-upload a shorter file. You can also open the trimmer on a clip that’s already within the limit if you just want to adjust it.
How long can my input video be when using Veo 3.1 Extend?
Veo 3.1 Extend extends your input video by 7 seconds. Your input video must be between 2-30 seconds long.
Generating Images and Videos
How do Unlimited generations work?
Unlimited gives you the freedom to create with selected models without worrying about credits. It’s designed to help you stay in your creative flow and iterate freely subject to reasonable use, daily usage thresholds per account, and the restrictions described below.
- No credit deductions: Generate on supported models without deducting from your credit balance.
- Continuous access: Keep creating on supported models even after your credits run out, unless reasonable use limits, daily usage thresholds, which may vary by selected AI model, service-performance protections, or misuse restrictions apply.
How do fast generations work?
Every month, you get a fixed number of top-priority fast generations with Unlimited models, depending on your plan. Once your fast generations run out, you still have Unlimited access, generations simply process at standard speed. If you continue to need fast generations after they have run out, toggle Unlimited off and use your credits. If you are on a team, the pool of fast generations are shared across the team.
How will I know how many fast generations I have left?
The number of Fast generations left will show when you hover over your credit display area.
How do parallel generations work?
Parallel generations determine how many image and video assets a single user can create at the same time. Your plan includes a set limit (e.g., 8, 10, or 12) on how many assets you can generate at once. This limit remains the same when you’re using Unlimited access.
How can I tell which model to use for my project?
The Model Picker helps you quickly find the most suitable AI model for your creative or production needs when generating videos, images, and voiceovers.
Instead of browsing all available models, you can search for a model by name or narrow down results using filters tailored to your workflow, preferences, and technical requirements.
You can apply a single filter or combine multiple filters to refine your results with greater precision.
To choose a model:
- Search by name or apply filters.
- Browse the available models.
- Click Use Model once you’ve found the best fit.
What format are the generated images?
The images are generated as JPEG.
What resolution are images generated at when generating images with an image reference?
The resolutions of generated images are model dependent. Read here to find out more about each model.
What format are generated videos?
The videos are generated as MP4.
What length of video can I make?
Depending on the model you choose, you can create videos that are 4, 5, 6, 8, 10, 12, or 15 seconds long (or up to 30 seconds with Seedance 2.5). If you need a longer video, try generating multiple short clips and stitching them together using your preferred video editing software.
What’s the difference between Kling 3.0 Motion Control and Kling 03 Video Edit?
Kling 3.0 Motion Control
What it does:
Takes motion from a reference video and applies it to a character/image.
What it’s best for:
- Making characters dance, fight, walk, act
- Recreating real performances
- Consistent, predictable motion
Example:
- Upload a TikTok dance → your AI character performs it exactly
Kling O3 Video Edit
What it does:
Edits or transforms an existing video using AI prompts.
Capabilities:
- Add/remove objects
- Change background or style
- Extend clips
- Replace elements
- Keep original motion but alter visuals
What it’s best for:
- Fixing or enhancing footage
- Turning real video into anime / cinematic style
Editing scenes without reshooting
Example:
- Take a real video → turn it into cyberpunk style
- Remove a person → replace with another character
What are some common tag functions I can use with Kling 3.0 Motion Control, Kling 03 Video Edit, Seedance 2.0, and Seedance 2.5?
First/Last Frame: Tell the AI exactly where to start the video
Example: Set @Image1 as the first frame
Character Consistency: Ensure the character looks like your reference throughout the clip
Example: Reference the appearance of @Image2
Motion Transfer: Takes the movement from a reference video and applies it to your character/subject
Example: Imitate the action of @Video1
Audio Sync: Align the video’s pacing and cuts to the beat of your uploaded music.
Example: Sync movement to the rhythm of @Audio1
Camera Control: Mimic specific lens work like a dolly zoom or a tracking shot
Example: Reference the camera movement from @Video1
The more specific you are with your @ tags (e.g., "@Image1 as character, @Video1 for motion), the more consistent and professional your final result will be. You can learn more about Kling 3.0 Motion, Kling 03 Video Edit, Seedance 2.0, and Seedance 2.5 here.
How can I see the name for tagging my references when using Kling 3.0 Motion Control, Kling 03 Video Edit, Seedance 2.0, and Seedance 2.5?
Hovering over the thumbnail of a given reference will indicate the name for tagging.
How do I enhance character consistency across multiple videos?
To improve consistency in your generated videos:
-
Start with an image – Use an Image to Video model so the AI has a clear visual reference to base the animation on.
-
Choose a model built for consistency – Some models are better at keeping characters, scenes, or motion stable across frames. Check the model guides to find one suited for consistent outputs.
- Refine your prompt – Make sure your prompt is clear and detailed. You can even run it through a tool like GPT to check for clarity or improve the phrasing before generating.
What image sizes can I create?
Different models support different aspect ratios. The available aspect ratios depend on the model you choose. To see which aspect ratios are supported, check the details for each model. If a model doesn’t support the selected aspect ratio, the generated image will match the size of your uploaded image.
Can I use a specific aspect ratio for generation?
Yes — but it depends on the model:
- If the model supports custom aspect ratios, the output will use the selected ratio.
- If not, the output will match the size of your uploaded image.
Check model details to see which aspect ratios are supported.
Can I generate images or videos in custom dimensions?
At this time, generation is limited to preset aspect ratios available per model, but you can crop or resize after download.
How long does it take to generate an image or video?
Image generation typically takes a few seconds, while videos may take slightly longer depending on length and complexity.
Can I increase the size of the prompt area?
Yes, you can increase or decrease the size of the prompt area by dragging the upper-right-hand corner of the prompt field
Can I revert an enhanced prompt back to the original or previous prompt?
Click the back button next to auto-prompt to revert to the previous prompt.
What is a negative prompt?
You can choose to include a Negative Prompt. A negative prompt is a way to tell the AI what you don’t want to appear in the final image or video.
While a normal prompt describes what you want to see, a negative prompt removes unwanted details, styles, or elements from the output.
Example:
Prompt: “A portrait of a woman in a studio, soft lighting.”
Negative prompt: “blurry, extra hands, distorted face, text.”
Using a negative prompt helps the AI avoid these issues, giving you a cleaner, more accurate result.
Do prompt areas have character limits?
There are limits on Kling (2500 characters), Seedance (5000), and Seedream 4.0 (5000) models, and image generation from text (2000). All other models support an unlimited number of characters in the prompt.
Can I generate a video in a language other than English?
If you are looking to generate videos in languages other than English, our Veo models support generations in all languages.
What error messages may I see when using auto-prompt?
Possible error messages you may see when using auto-prompt are:
- Your prompt is too short for auto-prompt. Add more detail and try again.
- Your prompt is too long for auto-prompt. Shorten it and try again.
- Your prompt may include restricted content. Edit it and try again.
- Auto-prompt failed. Please try again.
How does Model availability work with Unlimited?
Supported models may change from time to time, including for existing users. We may add, remove, replace, or limit models at our discretion. We will use reasonable efforts to communicate material changes as soon as reasonably practicable. Availability may vary between plans and offerings, including Artlist AI Tools and Studio, as some models may not be implemented or available in all offerings. Studio Unlimited excludes character and location generation, which may have separate limits or restrictions.
What is reasonable use when dealing with Unlimited models?
Unlimited is for individual use and is subject to reasonable use limits, including daily usage thresholds. It may not be used for automation, bots, scraping, account sharing, resale, bypassing limits, or abusive activity. We may apply restrictions where needed to protect the service. For details, please see our Terms of Use. The reset time for the fair use limit is midnight UTC.
Why can’t I find a model I used before?
We regularly retire older models as newer ones become available. As of September 2026, the following models have been removed from the model picker: Kling 1.6, Kling 2.1 Master, LTX 2.0 Pro, Sora 2, and Sora 2 Pro. Any content you already generated with these models stays available in your folders and library — you’ll just need to choose a different model for new generations.
Editing and using your generations
Can I edit or enhance generated images and videos?
Absolutely. Via your folders, open up a generation and click edit. This will pull your image or video back into the prompt area to start a new iteration.
Once your generations are downloaded, you can also edit your images and videos using any photo or video editing software.
Can I use my AI generated content after my subscription expires?
When you create AI-generated content with Artlist, it’s fully yours. You keep the rights to the generated outputs. You can use AI-generated outputs however you like, including for commercial purposes, even after your subscription expires, as long as you follow our license terms and avoid any forbidden uses.
Can I share my AI generations on social media?
You can freely share your generated content anywhere, including social media and portfolio platforms.
Can I use AI-generated content in commercial projects or client work?
Your generated images and videos can be used in any commercial context, including branded content, advertisements, social media, and client projects.
Can I use AI-generated visuals in products for resale (like posters, merch, or album covers)?
You can use AI-generated outputs however you like, including printed merchandise and advertisements, even after your subscription expires, as long as you follow our license terms and avoid any forbidden uses.
What happens to my rights to my AI-generated content if my subscription expires?
Any AI-content you generated while your subscription was active remains covered under the Artlist license. You can continue using it in existing and future projects, even after your subscription expires.
Note: AI-generated content downloaded from the Explore page in Artlist AI Tools or Artlist Catalog is considered Asset and is covered under the Asset license. If you complete and publish a project using these downloaded assets while your subscription is active, you retain the right to use them in that published project even after your subscription ends. However, once your subscription expires, these assets cannot be used in new projects, even if they were edited from the original.
Where can I see my previous generations?
You can see previous generations within your Folders.
Where can I see my downloaded generations?
When you click download, you can find your assets on your device or in 'Downloads' on the stock catalog.
Content Management and Privacy
Will my content be automatically backed up?
Yes, all content is backed up on our servers.
Is there a storage limit on the number of generations I can create?
No, there is no storage limit.
What happens when I delete a generation?
If you delete a creation, it is also removed from our server and cannot be recovered.
If my subscription expires, will I still have access to my outputs?
If your subscription expires, you won't be able to download the content you generated.
Only an active account can download content from the Artlist site.
Does anyone else have access to my generations?
Your creations are private and secure — no one else can view or use them without your permission.
However, in certain cases, our support team may access your content to help troubleshoot technical issues or provide assistance you’ve requested. This access is limited to authorized team members and is used solely for support purposes.
Does Artlist use my prompts, inputs, or outputs to train AI models?
Our AI services incorporate third-party AI models that are made available through our platform. These models are contractually prohibited from using your data to train or improve their AI models, unless we clearly indicate otherwise.
Some models, however, may use your inputs and outputs for AI training or model improvement purposes. Your decision to generate content using those models is optional and subject to your discretion.
Models that may use your data for AI training or improvement will be clearly identified with an information icon next to them, so you can review this indication before deciding whether to use the relevant model.
Do I need a license to complete and publish AI assets created by Artlist, that I downloaded or edited directly from Artlist?
Yes, you do. AI-generated content downloaded from the Explore page in Artlist AI Tools or Artlist Catalog is considered Asset and is covered under the Asset license. If you complete and publish a project using these downloaded assets while your subscription is active, you retain the right to use them in that published project even after your subscription ends. However, once your subscription expires, these assets cannot be used in new projects, even if they were edited from the original.
Need some more help? Head back to our Help Center.