AI Image Generator (CPU Version - Using Small Model)

Enter a short description or use voice input. The app uses OpenAI (if API key is provided) to create a detailed prompt, then generates an image using a small model (nota-ai/bk-sdm-tiny) on the CPU.

⚠️ Note: Using a small model for better compatibility on CPU. Generation should be faster than full Stable Diffusion, but quality/details may differ.

⏰ CPU generation can still take 1-5 minutes per image depending on load and model specifics.

Note: OpenAI API key found. Prompt will be enhanced using OpenAI.

Enter short description

Or record your idea (clears text box if used)

(Optional controls - Note: Their impact might vary on this small model)

Base Style

Negative Prompt (optional)

Guidance Scale (CFG)

1 10

Inference Steps (lower = faster but less detail, max 40)

5 40

Generated Prompt / Status

Generated Image