Basic Generation
Standard text-to-image and image-to-image generation using the same parameters as Flux.1 Dev.
Request Example
POST https://api.qubico.ai/v1/inference
Content-Type: application/json
x-api-key: YOUR_API_KEY
{
"model": "Qubico/flux1-dev-advanced",
"task_type": "txt2img",
"input": {
"prompt": "A serene mountain landscape at sunset, highly detailed, cinematic lighting",
"negative_prompt": "low quality, blurry, distorted, artifacts",
"width": 1024,
"height": 1024,
"guidance_scale": 4.0,
"num_inference_steps": 28,
"seed": 42
}
}Required Parameters
modelModel identifier. Must be "Qubico/flux1-dev-advanced"
task_typeTask type. Must be "txt2img" for text-to-image generation.
input.promptText description of the desired image. Maximum 1000 characters. Be specific and descriptive for best results.
Optional Parameters
input.negative_promptText description of what to avoid in the image. Maximum 500 characters.
input.widthOutput image width. Range: 256-2048, Default: 1024. Must be divisible by 8. width×height cannot exceed 4,194,304.
input.heightOutput image height. Range: 256-2048, Default: 1024. Must be divisible by 8. width×height cannot exceed 4,194,304.
input.guidance_scaleControls how closely the image follows the prompt. Range: 1.0-20.0, Default: 4.0
input.num_inference_stepsNumber of denoising steps. Range: 1-50, Default: 28. More steps = higher quality but slower generation.
input.seedRandom seed for reproducible results. Range: 0-2147483647. Use same seed for consistent outputs.
{
"id": "req_abc123def456",
"status": "completed",
"output": {
"images": [
{
"url": "https://cdn.qubico.ai/generated/image_123.png",
"width": 1024,
"height": 1024,
"seed": 42
}
]
},
"meta": {
"model": "Qubico/flux1-dev-advanced",
"task_type": "txt2img",
"credits_used": 3.0,
"generation_time": 2.8
}
}Prompt Writing
- • Be specific and descriptive
- • Include style keywords (e.g., "cinematic", "detailed")
- • Use negative prompts to avoid unwanted elements
- • Experiment with different prompt structures
Parameter Tuning
- • Start with default values and adjust gradually
- • Higher guidance_scale = stronger prompt adherence
- • More steps = higher quality but slower generation
- • Use seeds for reproducible results