Fluidfit brings together a range of AI models for image generation, image editing, video generation, and vector creation. Each model has its own strengths, supported inputs, and generation settings.
This guide will help you understand what each model is best suited for, choose the right one for your creative work, and explore the available options.
Choosing the right AI model
Use this quick guide to find a model based on what you want to create.
What do you need? | Models to try |
Speed and quick iterations | Z Image Turbo, Nano Banana 2 Lite, Hailuo 2.3 Fast |
High-quality, detailed images | Qwen Image 2 Pro, FLUX 2 Max, Seedream 5.0 Pro, FLUX 3 Image |
Branding, logos and vector assets | Recraft V4, Recraft V4 SVG |
Image generation and editing | Nano Banana 2.1, Grok Imagine Image 2.0, Wan 2.7 Image Pro, FLUX 3 Image |
Consistent characters in videos | Kling 3.0 Omni |
Cinematic video sequences | PixVerse V6, Seedance 2.0, Seedance 2.5 |
Animate an existing image | Grok Imagine Video, Gemini Omni 1.1 Flash |
Video generation using a reference video | Seedance 2.5 |
These are suggested starting points. The best model may vary depending on your prompt, input assets, and desired output.
Image generation and editing
Fluidfit offers several models for creating images from text prompts and editing existing images. Some models are optimized for speed and everyday creative work, while others focus on detailed visuals, professional-quality output, or more controlled editing.
GPT Image 2
Best for: Detailed, realistic images and complex creative prompts.
GPT Image 2 follows detailed instructions and creates professional-looking visuals, including complex scenes with multiple elements. It is particularly useful for marketing creatives, product mockups, advertising visuals, social media content, and concept art.
What makes it useful:
Strong understanding of detailed prompts.
Accurate text rendering within images.
Handles complex scenes and multiple visual elements.
Produces realistic and polished creative assets.
Settings available:
Quality
Aspect Ratio
Output Format
Output Compression
Moderation
Prompt tip: Describe the subject, environment, lighting, visual style, and composition clearly. Specific prompts generally produce better results.
Seedream 5.0 Lite
Best for: Fast, clean, and reliable image generation.
Seedream 5.0 Lite is a good choice when you want consistent results without spending too much time refining prompts or retrying generations.
What makes it useful:
Faster generation for quick iterations.
More consistent outputs with fewer retries.
Cleaner compositions and fewer unwanted visual artifacts.
Works well with simple prompts.
Settings available:
Size: Choose a higher size, such as 2K, for sharper images.
Aspect Ratio: Select the format that suits your output.
Output Format: PNG or JPG.
Available in Generate Image and Edit Image.
Seedream 5.0 Pro
Best for: High-quality image generation and editing with reference guidance.
Seedream 5.0 Pro lets you generate new images and refine existing ones using the same model. It is useful when you need detailed output or want to maintain consistency across edits.
What makes it useful:
Generate and edit images without switching models.
Use reference images to maintain character, product, or style consistency.
Produces sharp, high-resolution output up to 2K.
Settings available:
Aspect Ratio: Standard options, including Match Input Image when using AI Edit.
Size: 1K or 2K.
Output Format: PNG or JPEG.
Available in Generate Image and Edit Image.
Nano Banana 2 Lite
Best for: Fast image generation, quick iterations, and accurate text rendering.
Nano Banana 2 Lite is designed for speed and affordability. It works well for exploring ideas, generating variations, and creating images with readable text.
What makes it useful:
Fast generation for high-volume creative work.
Accurate text rendering within images.
Supports both image generation and editing.
Reference images can guide edits and help match a desired look.
Settings available:
Aspect Ratio: Multiple standard ratios, including portrait, landscape, square, and ultrawide formats.
Output Format: JPG or PNG.
Match Input Image is available when using AI Edit.
Prompt tip: Keep prompts simple and direct. When editing text within an image, specify exactly what the new text should say.
Nano Banana 2.1
Best for: Image generation and editing with control over image quality and output.
Nano Banana 2.1 is available in both Generate Image and Edit Image. It offers additional controls for image quality, aspect ratio, and output format.
What makes it useful:
Supports image generation and editing.
Offers multiple output resolutions.
Provides control over the level of thinking used during generation.
Settings available:
Aspect Ratio: Multiple standard aspect ratios.
Resolution: 1K, 2K, or 4K.
Thinking Level: Minimal, Medium, or High.
Output Format: JPG or PNG.
Grok Imagine Image
Best for: Creating and editing images using natural language prompts.
Grok Imagine Image lets you create new visuals or make targeted changes to existing images without starting over.
What makes it useful:
Available in Generate Image and Edit Image.
Edit images by describing the changes you want.
Useful for replacing objects, updating colours, changing styles, extending scenes, and creating variations.
Settings available:
Aspect Ratio: Auto and multiple standard ratios.
Resolution: Available resolution options are displayed in the model settings.
Grok Imagine Image 2.0
Best for: Image generation and editing.
Grok Imagine Image 2.0 is available for both creating new images and editing an existing image.
What makes it useful:
Supports Generate Image and Edit Image.
Allows an existing image to be used as an input for editing.
Settings available:
Aspect Ratio: Auto and multiple standard ratios.
Resolution: Auto, 1K, and 2K.
Qwen Image 2 Pro
Best for: Detailed, photorealistic images and complex creative prompts.
When image detail and prompt understanding matter, Qwen Image 2 Pro is a strong choice. It is suitable for product photography, illustrations, advertising creatives, and scenes with multiple subjects.
What makes it useful:
Detailed, high-quality image output.
Strong prompt understanding.
Suitable for complex scenes and multiple subjects.
Available in Generate Image and Edit Image.
Qwen Image 2
Best for: Fast creative exploration and everyday design work.
Qwen Image 2 balances quality and speed, making it useful for brainstorming, creating social media content, exploring product concepts, and quickly iterating on designs.
FLUX 2 Max
Best for: Premium marketing assets and professional-quality visuals.
FLUX 2 Max produces rich, polished images with strong prompt understanding. It is a good choice for high-end marketing campaigns, realistic portraits, product photography, and commercial creatives where visual quality is the priority.
Wan 2.7
Best for: Everyday image generation and natural image editing.
Wan 2.7 offers a balance of quality and flexibility for marketing creatives, concept art, product visuals, and day-to-day creative work.
Available in Generate Image and Edit Image.
Wan 2.7 Image Pro
Best for: Image generation and editing with control over output size and seed.
Settings available:
Size: 1K, 2K, or 4K, along with supported preset dimensions.
Seed: Random or a specified seed value.
FLUX 3 Image
Best for: High-resolution image generation and editing.
FLUX 3 Image is available for both Generate Image and Edit Image.
Settings available:
Aspect Ratio: Multiple standard and creative aspect ratios.
Resolution: 1K, 1.5K, 2K, or 4K.
Output Format: JPG, PNG, or WebP.
Recraft V4
Best for: Marketing creatives, branding, and everyday design work.
Recraft V4 produces clean, visually appealing results for social media graphics, advertisements, product visuals, posters, and promotional content.
Recraft V4 Pro
Best for: Professional-quality marketing assets and premium design work.
Recraft V4 Pro builds on the strengths of Recraft V4 with improved detail, stronger prompt understanding, and more refined outputs. Choose it for client-ready creatives and commercial designs.
Hunyuan Image 3
Best for: Rich artistic visuals and detailed illustrations.
Hunyuan Image 3 is suited to expressive artwork, realistic imagery, concept art, storytelling, posters, and character designs.
Z Image Turbo
Best for: Fast image generation and rapid creative iteration.
Z Image Turbo is designed for speed. Use it to explore multiple ideas, prototype concepts, and quickly generate visual variations.
Higgsfield Soul 2.0
Best for: Image generation.
Higgsfield Soul 2.0 is available in Generate Image.
Settings available:
Resolution: 720p and 1080p.
Vector generation
Vector graphics are made from shapes and paths rather than pixels. Unlike regular images, they can be scaled up or down without losing quality, making them useful for branding, printing, presentations, websites, and marketing materials.
Recraft V4 SVG
Best for: Logos, icons, illustrations, and everyday vector graphics.
Recraft V4 SVG generates clean, scalable vector artwork for digital and print use.
It is useful for:
Brand assets and logos.
UI icons and interface elements.
Infographics and stickers.
Simple illustrations.
Recraft V4 SVG Pro
Best for: Professional branding and production-ready vector artwork.
Recraft V4 SVG Pro offers more refined vector output and stronger prompt understanding. Choose it when creating polished logos, detailed illustrations, or commercial assets.
Where to find these models: Open the Generate Vector tool and select your preferred Recraft SVG model.
Video generation
Fluidfit offers video models for creating videos from text prompts and visual inputs. Depending on the model, you can guide the output using first frames, last frames, reference images, or reference videos.
Kling 3.0 Omni
Best for: Guided video generation, character consistency, and multi-scene storytelling.
Kling 3.0 Omni supports text prompts, images, or a combination of both. It is useful when you need more control over how a video develops.
What makes it useful:
Supports text-only, image-only, and text-plus-image workflows.
Use reference images for characters, products, styles, and scenes.
Supports multi-prompt workflows to guide different parts of a video.
Useful for scene transitions, story progression, and multiple actions in one video.
Settings available:
Aspect Ratio.
Duration.
Mode: Standard (720p), Pro (1080p), or 4K.
Generate Audio.
Seedance 2.0
Best for: Smooth motion, scene consistency, and creative video generation.
Seedance 2.0 is designed to produce cleaner frames, smoother motion, and polished video output. It works well for advertisements, product clips, concept videos, social media content, and storytelling.
Supported inputs:
Text prompt.
First Frame.
Last Frame (for Generate Video).
Settings available:
Duration.
Resolution.
Aspect Ratio.
Generate Audio.
Seed value.
Tip: Start with shorter videos at 720p while testing prompts. Increase the quality settings once you're happy with the result.
Seedance 2.5
Best for: Video generation using reference videos and guided creative inputs.
Seedance 2.5 supports reference-video-based generation, giving you another way to guide the resulting video using existing footage.
Settings available:
Resolution: 480p or 720p.
Aspect Ratio: Supported options include 16:9, 4:3, 3:4, 9:16, and other displayed presets.
Duration: Auto or a manually selected duration from 5 to 30 seconds.
Seed: Random or a specified value.
Generate Audio: Optional.
Duration tip: Set Duration to Auto to let the model select an appropriate video length.
PixVerse V5
Best for: Everyday video generation and faster creative exploration.
PixVerse V5 is a reliable option for general-purpose video creation and existing workflows that use PixVerse.
Supported inputs:
Text prompt.
Image.
First Frame.
Last Frame.
PixVerse V6
Best for: Cinematic sequences, polished marketing content, and multi-shot storytelling.
PixVerse V6 offers improved prompt understanding, more natural motion, higher visual fidelity, and better scene consistency.
What makes it useful:
More natural motion and richer cinematic results.
Improved scene consistency.
Multi-shot video generation for connected scenes.
Useful for product showcases, commercials, and narrative content.
Supported inputs:
Text prompt.
Image.
First Frame.
Last Frame.
Grok Imagine Video
Best for: Image-to-video generation, cinematic motion, and video editing.
Grok Imagine Video can turn a text prompt or an image into a dynamic video. It also supports editing existing videos using natural language instructions.
What makes it useful:
Bring still images to life with realistic movement.
Create smooth camera movements and cinematic scenes.
Generate synchronized background audio and sound effects.
Create variations from the same idea or source media.
Edit existing videos by describing the changes you want.
Available in Generate Video and Edit Video.
Grok Imagine 1.5 Video
Best for: Natural-looking motion, cinematic scenes, and creative storytelling.
Grok Imagine 1.5 Video creates videos from text prompts or images. It is suitable for marketing content, product showcases, promotional videos, character-driven content, and social media creatives.
What makes it useful:
Smooth, natural-looking motion.
Strong visual quality for cinematic scenes.
Image-to-video generation.
Useful for product showcases and character-driven content.
Available in Generate Video.
Hailuo 2.3
Best for: High-quality video generation and polished creative work.
Hailuo 2.3 is designed to produce smooth, cinematic videos with natural motion. It works well for product showcases, marketing content, character animations, and social media videos.
Supported inputs:
Text prompt.
First Frame image (optional).
Settings available:
Duration: 6 or 10 seconds.
Resolution: 768p or 1080p.
Prompt Optimizer: Optional, to automatically enhance prompts.
Duration limitation: 10-second videos are available only at 768p. At 1080p, the maximum supported duration is 6 seconds.
Hailuo 2.3 Fast
Best for: Faster video generation and rapid iterations.
Hailuo 2.3 Fast offers a faster alternative for exploring creative ideas, testing prompts, and generating variations.
Settings available for both Hailuo models:
Duration: 6 or 10 seconds.
Resolution: 768p or 1080p.
Prompt Optimizer: Optional, to automatically enhance prompts.
Duration limitation: 10-second videos are available only at 768p. At 1080p, the maximum supported duration is 6 seconds.
Choose Hailuo 2.3 when output quality is the priority, or Hailuo 2.3 Fast when speed and rapid experimentation matter more.
Gemini Omni 1.1 Flash
Best for: Turning prompts, sketches, and reference images into videos.
Gemini Omni Flash can generate videos from text alone or use visual inputs to guide the output. It is useful for bringing a sketch or reference image to life.
What makes it useful:
Turn a rough sketch into realistic motion.
Animate a single reference image.
Generate videos from text prompts alone.
Combine multiple visual inputs to guide the result.
Supported inputs:
Text.
First Frame.
Last Frame.
Reference Images.
Settings available:
Aspect Ratio: 16:9 or 9:16.
Resolution: Available options are displayed in the model settings.
HappyHorse 1.0
Best for: AI video generation.
HappyHorse 1.0 is available in Fluidfit's video generation model selection.
A few things to keep in mind
Not every model supports the same inputs or generation settings. Check the options shown in the selected model's settings panel.
First Frame, Last Frame, reference images, and reference videos are available only on supported models.
Resolution, duration, and aspect ratio options can vary by model.
Higher resolutions and longer videos may take more time to generate.
The best model depends on your creative goal, prompt, input assets, and desired output.
For details about models that are no longer available, see Deprecated Models.