Overview
Open Canva provides access to 25+ image models and 20+ video models from leading AI providers. Each model has unique strengths — understanding them helps you get the best results for your use case.Image models
Flagship models
| Model | Best for | Notes |
|---|---|---|
| Flux Schnell | Fast photorealistic images | Quick iteration, good default |
| Flux 2 Max | Premium photorealism | Highest-quality Flux output |
| Flux 2 Flex | Flexible creative output | Good balance of speed and quality |
| GPT Image 1.5 | General-purpose, versatile | OpenAI’s latest image model |
| Ideogram V3 | Text in images, typography | Best for designs that include words |
| Imagen 4 Preview | Google’s flagship image model | High fidelity and prompt adherence |
| Seedream V5 Lite | Fast, high-quality generation | Good for rapid exploration |
| Seedream V4.5 | Detailed creative imagery | Strong style control |
Specialized models
| Model | Best for | Notes |
|---|---|---|
| Grok Imagine | Creative and artistic output | xAI’s image generation |
| Kling O1 Image | Cinematic, high-detail images | Strong composition |
| Nano Banana | Fast affordable generation | Budget-friendly option |
| Nano Banana Pro | Higher-quality Nano output | Better detail than base |
| Nano Banana 2 | Latest Nano generation | Improved quality and speed |
| OneReward | Reward-optimized output | Trained on human preferences |
| Qwen Image 2 | Multi-style generation | Versatile Chinese AI model |
| Qwen Image 2 Pro | Premium Qwen output | Higher detail and fidelity |
| Runway Gen-4 Image | Cinematic, film-like stills | Runway’s latest model |
| Wan v2.6 | General-purpose | Versatile across styles |
Video models
Open Canva supports both text-to-video (T2V) and image-to-video (I2V) generation:| Model | T2V | I2V | Notes |
|---|---|---|---|
| Kling 2.6 Pro | Yes | Yes | High-quality cinematic video |
| Kling V3 Standard | Yes | Yes | Balanced quality and speed |
| Kling V3 Pro | Yes | Yes | Premium Kling output |
| MiniMax Hailuo 2.3 | Yes | Yes | Fast, good motion quality |
| Google Veo 3.1 | Yes | Yes | Google’s latest video model |
| Wan 2.6 | Yes | Yes | Versatile general-purpose video |
| Grok Imagine Video | Yes | Yes | xAI video generation |
| Kandinsky 5 Pro | Yes | Yes | Artistic video styles |
| Kandinsky 5 Distill | Yes | — | Fast artistic video |
| LTX-2 19B | Yes | Yes | Open-source high-quality |
| LTX 2.3 | Yes | Yes | Fast LTX variant |
| Decart Lucy 14B | — | Yes | Image-to-video specialist |
Video Generation Guide
Learn about video generation modes, duration limits, and download formats.
Choosing the right model
Identify your goal
- Photorealistic photo: Flux 2 Max, GPT Image 1.5, Imagen 4
- Design with text: Ideogram V3
- Fast iteration: Flux Schnell, Seedream V5 Lite, Nano Banana
- Cinematic look: Runway Gen-4, Kling O1
- Video content: Kling V3 Pro, Veo 3.1, Hailuo 2.3
Consider speed vs. quality
Faster models (Flux Schnell, Nano Banana) are great for exploring ideas. Premium models (Flux 2 Max, Imagen 4) produce higher fidelity but take longer.
Model compatibility
Not all models support every aspect ratio or feature. The app automatically:- Grays out models that don’t support your selected aspect ratio
- Shows which modes (T2V, I2V) each video model supports
- Validates settings before generation
Related guides
Multi-Model Generation
Compare results across multiple models.
Settings Guide
Configure aspect ratio and other parameters.
Video Generation
Generate videos with AI models.
Dashboard Analytics
Track which models you use most.