flux_text_to_image
Generate images from text prompts using local Flux models. Auto-detects installed models to produce 1024x1024 images.
Instructions
Generate images using Flux (local model).
All parameters except prompt are optional with sensible defaults. Auto-detects installed Flux model (GGUF or safetensors) and CLIP encoders. Produces high-quality 1024x1024 images by default.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| seed | No | Random seed. Default: -1 (random). | |
| steps | No | Sampling steps. Default: 20. Typical: 15-30. | |
| width | No | Image width in pixels. Default: 1024. Must be multiple of 16. | |
| height | No | Image height in pixels. Default: 1024. Must be multiple of 16. | |
| prompt | Yes | Text description of the desired image. | |
| guidance | No | Prompt guidance strength. Default: 3.5. Typical: 2.0-5.0. Note: Flux embeds guidance in conditioning, not in CFG. | |
| vae_name | No | VAE model. Default: "ae.safetensors". Use list_models(folder="vae") to see options. | ae.safetensors |
| batch_size | No | Number of images to generate. Default: 1. | |
| clip_name1 | No | First CLIP model (clip_l). Default: "" (auto-detects). | |
| clip_name2 | No | Second CLIP model (t5xxl). Default: "" (auto-detects). | |
| diffusion_model | No | Flux model filename. Default: "" (auto-detects). Supports both safetensors and GGUF formats. Use list_models(folder="diffusion_models") to see options. |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||