The AI models behind Melon Cut Design
Melon Cut Design brings several AI models into one studio, on one credit balance. Pictures are made with GPT Image, videos with the video model you pick - or with Auto, which picks one for you - and vectors with Recraft.
- video models in the picker
- 7
- longest clip, on the longest models
- 30 s
- credit balance for every model
- 1
AI models, step by step
Pick the kind of chat
Start an Image, Video, Vector or Brand chat. Each kind uses the models made for that job.
Choose a video model, or Auto
In a video chat, pick a model from the list or leave it on Auto, which chooses one for your prompt and says why.
See what was used
Open a result's details to see the models used, the settings, the prompt and what it cost.
Pictures: GPT Image
Every picture Melon Cut generates is made with GPT Image. There is no model menu for pictures: with one model you learn how it responds to your prompts, and the studio's settings - format, shape, quality and brand - do the rest.
Video: a model for each kind of clip
Video models differ in how long a clip they make and which shapes they support, so the length and shape you can choose depend on the model you pick. Auto takes that decision off your hands: it picks a model for your prompt and tells you why.
- Kling 4.0: 3 to 30 seconds
- Wan 3.0: 2 to 30 seconds
- Seedance 2.5: 4 to 30 seconds
- Veo 3.1: 4, 6 or 8 seconds
- Gemini Omni Flash: 3 to 10 seconds
- Grok Imagine: 1 to 15 seconds
- Happy Horse: changes to an existing video
Vectors: Recraft
Vector graphics and Convert to vector use Recraft vector models. They return real, editable SVG - paths and shapes you can open in a vector editor - rather than pixels.
Every model on every plan
Plans differ only in their monthly credits, so every model is there whichever plan you are on. A picture uses 10 credits, a vector 10, and a video 60 for every started 5 seconds.
AI models: the details
- Pictures
- GPT Image
- Video
- Kling 4.0, Wan 3.0, Seedance 2.5, Veo 3.1, Gemini Omni Flash, Grok Imagine
- Video edits
- Happy Horse
- Vectors
- Recraft vector models
- Video shapes
- 16:9, 9:16, 1:1, 21:9, 4:3 or 3:4, depending on the model
- Choosing
- Pick a video model, or let Auto choose
Can I choose the model for pictures?
No. Pictures are always made with GPT Image. Model choice is offered for video, where models differ in length and shape.
What does Auto do?
In a video chat, Auto picks a video model for your prompt and tells you why it chose it. You can pick a specific model instead whenever you prefer.
Which model makes the longest videos?
Kling 4.0, Wan 3.0 and Seedance 2.5 make clips of up to 30 seconds. Grok Imagine goes up to 15, Gemini Omni Flash up to 10, and Veo 3.1 makes 4, 6 or 8 seconds.
Works well with
- Create images AI image generator Text to image, up to 4 pictures per prompt.
- Video AI video generator Text to video, up to 30 seconds, in the shape you need.
- Vectors AI vector generator Real, editable SVG from a text description.
- Video AI video editor Change an MP4 or MOV of up to 60 seconds with a prompt.
People also ask
- AI images Which AI models does Melon Cut use? Like most AI design tools, Melon Cut uses different models for different jobs. Melon Cut Design makes pictures with GPT Image, with no model to choose. For video you can pick Kling 4.0, Wan 3.0, Seedance 2.5, Veo 3.1, Gemini Omni Flash or Grok Imagine, or let Auto pick and explain why; Happy Horse handles video edits. Vectors use Recraft's vector models.
- Video How long can AI videos be? AI video models generate short clips, from a few seconds to about half a minute at a time. In Melon Cut Design the longest-running models make up to 30 seconds: Kling 4.0 makes 3 to 30, Wan 3.0 2 to 30 and Seedance 2.5 4 to 30. Veo 3.1 makes 4, 6 or 8 seconds. For a longer video, join several clips in a video editor.
Try it on your next post
GPT Image, seven video models and Recraft vectors in one place.