Using AI Models
Browse the available AI models, understand their capabilities, and connect them in workflows.
Wireflow integrates dozens of AI models for image generation, video creation, audio synthesis, and more. This guide helps you find and use the right model for your project.
Browsing Models
When adding a node to a workflow, the node picker shows all available models organized by category:
- Image Generation — Nano Banana, GPT Image, Flux, Seedream, Imagen, Ideogram, Recraft, Qwen Image
- Video Generation — Kling, Seedance, Veo, Wan, MiniMax Hailuo, LTX, Pixverse
- Audio — ElevenLabs text-to-speech and voice swap, music generation, sound effects
- Text / LLM — GPT, Claude, Gemini, Grok for text processing
- Utility — Face swap, upscaling, background removal
The full list with per-node credit estimates is at All Nodes, generated from the same catalog the editor uses. If a model is not on that page, it is not available here.
Each model card shows its name, a brief description, and credit cost per run.
Model Inputs and Outputs
Every model node has typed inputs and outputs:
Common Inputs
| Input | Description |
|---|---|
| Prompt | Text instruction describing what to generate |
| Image | An input image (for image-to-video, style transfer, etc.) |
| Negative Prompt | What to avoid in the output |
| Aspect Ratio | Output dimensions like 16:9, 1:1, 9:16 |
Common Outputs
| Output | Description |
|---|---|
| Image | A generated image (PNG/JPEG) |
| Video | A generated video (MP4) |
| Audio | Generated audio (MP3/WAV) |
| Text | Generated or processed text |
Connecting Models
You can chain models by connecting the output of one to the input of another. Wireflow automatically validates that connections are compatible.
Example pipeline:
- Text Input → provides the prompt
- Flux Image → generates an image from the prompt
- Kling Video → animates the image into a video
Model Settings
Click a model node to see its settings panel. Common options:
- Quality / Steps — More steps = higher quality, more credits
- Guidance Scale — How closely the model follows your prompt
- Seed — Fix a seed for reproducible outputs
- Duration — For video/audio models, the output length
Credit Costs
Different models cost different amounts of credits per run. More powerful models and longer outputs cost more. See Credits & Billing for details on purchasing credits.