> ## Documentation Index
> Fetch the complete documentation index at: https://docs.imagine.art/llms.txt
> Use this file to discover all available pages before exploring further.
# AI Models in Workflows
> Browse all image, video, and text AI models available on the Workflows canvas.
ImagineArt Workflows gives you access to 50+ AI models across three categories — image, video, and text — all available from the same canvas. Each node type (Image, Video, Audio) has a built-in **model selector**: drop a single node onto the canvas, pick any model from the dropdown, and the node's available parameters update dynamically to match that model's capabilities. You no longer need separate nodes for each model — one node handles them all.
Model availability may change as new models are added. The lists below reflect the current model catalog.
Image models are available in the [Generate Image](/workflows/understanding-nodes#generate-image), [Edit Image](/workflows/understanding-nodes#edit-image), and [Upscale Image](/workflows/understanding-nodes#upscale-image) nodes.
### Generation and editing models
| Model | Modes | Description |
| ------------------------------ | -------------- | ---------------------------------------------------------------------------- |
| **Nano Banana Pro** | Generate, Edit | Google's state-of-the-art image generation and editing model |
| **Seedream V4.5** | Generate, Edit | ByteDance Seedream 4.5 for text-to-image and image editing |
| **Nano Banana** | Generate, Edit | Fast-paced, dynamic rendering model for generation and editing |
| **Seedream V4** | Generate, Edit | ByteDance Seedream 4 for generating high-quality images |
| **Flux 2** | Generate, Edit | Great aesthetics and prompt adherence for text-to-image generation |
| **Ideogram Character** | Generate, Edit | Character-focused Ideogram generation and editing |
| **ChatGPT Image** | Generate, Edit | OpenAI's most advanced image model for realistic images |
| **Flux Pro Kontext** | Generate, Edit | Flux 1.0 Pro Kontext with text and reference image inputs for editing |
| **Flux Pro Kontext Max** | Generate, Edit | Premium editing with stronger prompt adherence and typography |
| **Flux Pro Kontext Max Multi** | Generate, Edit | Experimental multi-reference image editing for complex compositions |
| **Qwen Image** | Generate, Edit | Text-to-image with parallel CFG, LoRA, and turbo acceleration |
| **Flux Pro 1.1 Ultra** | Generate, Edit | Advanced version of Flux Pro 1.1 for detailed, realistic scenes |
| **Ideogram V3** | Generate, Edit | Advanced typography handling for intricate text elements and design visuals |
| **ImagineArt 1.5** | Generate | Great aesthetics and prompt adherence for high-quality images |
| **Z Image Turbo** | Generate | Latest image generation model from Alibaba with high-speed results |
| **Seedream V3** | Generate | ByteDance Seedream 3.0 for text-to-image generation |
| **Dreamina 3.1** | Generate | ByteDance Dreamina 3.1 with rich styles and enhanced prompt features |
| **Flux Pro 1.1** | Generate | High-quality text-to-image offering realistic and detailed outputs |
| **Flux Dev** | Generate | Flexible and customizable model ideal for experimental and conceptual design |
| **Recaraft V3** | Generate | Text-to-image designed for vector and typography-friendly outputs |
| **Minimax Image 01** | Generate | Fast, efficient text-to-image for quick, high-quality generation |
| **Ideogram V2 Turbo** | Generate | Fast, high-efficiency model for typography-heavy content |
### Upscale models
Used in the [Upscale Image](/workflows/understanding-nodes#upscale-image) node.
| Model | Description |
| ----------------------------- | ---------------------------------------------------------------------------- |
| **ImagineArt Subtle** | Enhances resolution with subtle refinement and natural detail |
| **Topaz Upscale** | Industry-leading upscaling technology enhancing texture while reducing noise |
| **ImagineArt Creative** | Creative upscaling with artistic detail and visual style enhancement |
| **FreePik Upscaler Creative** | AI-driven upscaling introducing stylistic detail and artistic enhancement |
| **FreePik Precision v1** | Focuses on image fidelity, sharpness, and natural color preservation |
| **FreePik Precision v2** | Advanced precision upscaling with adaptive content-aware algorithms |
For maximum multi-reference editing capability, use **Nano Banana Pro**, **Nano Banana 2**, or **Seedream V4.5**—these models support uploading 10+ reference images in the Edit Image node.
Video models power the [Generate Video](/workflows/understanding-nodes#generate-video), [Edit Video](/workflows/understanding-nodes#edit-video), [Extend Video](/workflows/understanding-nodes#extend-video), [Lipsync](/workflows/understanding-nodes#lipsync), [Motion Transfer](/workflows/understanding-nodes#motion-transfer), and [Upscale Video](/workflows/understanding-nodes#upscale-video) nodes.
### Generation models
| Model | Description |
| ----------------------- | ------------------------------------------------------------------------- |
| **Seedance 1.5 Pro** | Stable, fluid animations for high-end video production |
| **Seedance Pro Fast** | High-speed ByteDance model with flexible camera controls |
| **Wan 2.6** | Alibaba's cinematic multi-shot video AI with audio |
| **Kling 2.6 Pro** | Advanced AI video generation with detailed text/image-to-video creation |
| **Kling Omni** | Versatile video generator with multi-frame reference control |
| **Veo 3.1** | Google's high-quality cinematic video with audio generation |
| **Veo 3.1 Fast** | Rapid, cost-effective Google video with high-fidelity output |
| **Sora 2 Pro** | OpenAI's advanced physics-based cinematic video generation |
| **Pixverse v5.5** | Pro-level cinematic video with advanced motion brush control |
| **Hailuo 2.3 Pro** | Premium 1080p cinematic video with ultra-realistic physical dynamics |
| **Veo 3 Fast** | Rapid generation for quick creative iterations |
| **Sora 2** | Versatile world simulation with high-quality physics |
| **Wan 2.2 Turbo** | High-speed generation with advanced storytelling controls |
| **Wan 2.5** | Alibaba's realistic video model with native audio |
| **Kling 2.5 Pro Turbo** | Ultra-fluid motion with precise prompt control |
| **Hailuo 2.3 SD** | Balanced high-speed video with enhanced physics and micro-expressions |
| **Kling 2.1 Pro** | Professional motion quality for stable visuals |
| **Seedance 1.0 Pro** | ByteDance's high-quality text/image to video with flexible camera control |
| **Pixverse v5** | Hyper-realistic textures and improved consistency across frames |
| **Veo 2** | Cinematic videos with professional camera controls |
| **Kling 2.1 SD** | Balanced performance for high-speed video generation |
| **Pixverse v4.5** | Viral social media video with cinematic motion |
| **Pixverse v4** | Creative video generation with versatile cinematic camera controls |
| **Kling 1.6 SD** | Optimized for cinematic motion and complex narrative consistency |
| **Kling 1.6 Pro** | Maximum visual fidelity and longer narrative sequences |
| **Hailuo 02 Pro** | Handles complex physical interactions—object collisions, fluid dynamics |
| **Hailuo 02 Standard** | Optimized for rapid iteration and creative experimentation |
| **Seedance 1.0 Lite** | Fast and efficient ByteDance generation with flexible aspect ratios |
| **Decart Lucy 14B** | Lightning-fast image-to-video generation with cinematic motion |
### Upscale models
| Model | Description |
| ---------------------- | --------------------------------------------------------------- |
| **ImagineArt Upscale** | Upscales videos while maintaining quality and enhancing details |
| **Topaz Upscale** | Enhances video resolution with advanced algorithms for clarity |
### Lipsync models
| Model | Description |
| -------------------------------- | ------------------------------------------------------------------ |
| **Kling AI Avatar Pro i2v** | High-fidelity talking avatar with expressive emotion |
| **Kling AI Avatar Standard i2v** | Efficient, stable talking avatar for everyday use |
| **Omnihuman v1.5** | ByteDance's realistic full-body motion and lip-sync model |
| **Kling AI Avatar Pro** | Professional-grade lip-sync with detailed facial micro-expressions |
| **Kling AI Avatar Standard** | Reliable audio-driven lip-sync for static portraits |
| **Infinitalk Audio** | Video dubbing with precise lip synchronization |
### Motion transfer models
| Model | Description |
| ------------------- | ------------------------------------------------------------ |
| **Wan 2.2 Replace** | Transfers video motion to static character images |
| **Wan 2.2 Move** | Animates static characters using video motion reference |
| **Runway Act Two** | Advanced character animation with video performance transfer |
| **MoonValley** | Creative video transformation with style and motion control |
### Extend models
| Model | Description |
| ------------------------ | ------------------------------------------------------------- |
| **Pixverse Extend** | High-quality video extensions with seamless visual continuity |
| **Pixverse Extend Fast** | Rapid video lengthening with optimized processing speed |
Text models power the [AI Copilot](/workflows/understanding-nodes#ai-copilot) node. These are large language models (LLMs) from leading AI providers, available for generating, analyzing, and transforming text within your workflows.
| Model | Use cases |
| -------------------------- | --------------------------------------------------------------------------------------- |
| **Gemini 3.0 Pro Preview** | Multimodal powerhouse for complex reasoning, creative text, image, and video generation |
| **Gemini 2.5 Pro** | Google's model for deep text analysis, multi-step reasoning, and complex tasks |
| **GPT-5.1** | OpenAI's top-tier model for high-performance problem-solving and creative text |
| **GPT-5 Mini** | Optimized for fast response times; ideal for high-volume text generation |
| **GPT-4o** | Omni model with speed and multimodal capabilities for text, voice, and video tasks |
| **Claude Sonnet 4.5** | Balanced workhorse for content generation, analysis, and general-purpose workflows |
| **Grok 4** | Distinctive personality, real-time knowledge integration, and dialogue |
| **GPT-5 Nano** | Small, efficient model for basic text processing and simple tasks |
| **GPT-4o Mini** | Fast, cost-effective multimodal model for quick text generation |
| **Gemini 2.5 Flash** | Speed-optimized for high-quality multimodal tasks; ideal for large text content |
| **Gemini 2.0 Flash** | Efficient model for common tasks like blog posts, social media, and scripts |
| **Gemini 2.5 Flash Lite** | Compact model for rapid, low-latency text responses in real-time applications |
| **Claude Opus 4.1** | Ideal for complex reasoning, creative writing, and detailed content generation |
| **Claude Haiku 4.5** | Fast, cost-effective model for quick tasks: dialogue, captions, and summaries |
For workflows that require analyzing images or videos as inputs to the AI Copilot node, choose a multimodal model such as **Gemini 3.0 Pro Preview**, **GPT-4o**, or **Gemini 2.5 Flash**.