BFL
This page is auto-generated from model configurations. Last updated: 2026-09-01.
This reference lists all available BFL image upscale models and their parameters. Use these parameter names when calling the Generation API.
Scenario Flux Upscale
Section titled “Scenario Flux Upscale”Upscale and enhance images with FLUX, adding realistic detail up to 4K with control over structure fidelity, creativity, and style.
Model ID: model_sc-upscale-flux
Capabilities: img2img
LLM Markdown: https://app.scenario.com/api/models/model_sc-upscale-flux/markdown
| Parameter | Type | Required | Default | Min | Max | Allowed Values | Description |
|---|---|---|---|---|---|---|---|
image | file | Yes | - | - | - | - | The image you want to upscale. |
upscaleFactor | number | No | 2 | 2 | 8 | - | How much to enlarge the image. The model adds real new detail up to 4096x4096px (for example, a 2048x2048 image at 2x); beyond that size it enlarges without inventing new detail. |
preset | string | No | balanced | - | - | precise, balanced, creative | A quick starting point that sets Structure Fidelity and Creativity together. Precise stays closest to the original, Creative adds the most new detail, Balanced sits between. |
loras | model_array | No | “ | - | - | - | Optional add-on style models (LoRAs) to apply while upscaling, by ID or URL. Add up to 10. |
lorasScale | number_array | No | “ | - | - | - | How strongly each LoRA is applied, in the same order as the models above. |
prompt | string | No | - | - | - | - | Optional text describing the image or the detail you want emphasized as it’s enhanced. |
imagePrompt | file_array | No | - | - | - | - | Optional images to guide the visual style of the result. Add up to 4. |
imagePromptStrength | number | No | 0.6 | 0 | 1 | - | How closely the result follows your style images. Higher values match them more tightly. |
controlnetConditioningScale | number | No | 0.4 | 0 | 1 | - | How closely the result preserves the original image’s structure. Higher values stay truer to the original layout and shapes. |
strength | number | No | 0.4 | 0 | 1 | - | How much new, imaginative detail the model adds. Higher values invent more; lower values stay faithful to the source. |
numInferenceSteps | number | No | 28 | 1 | 50 | - | How many refinement passes the model makes. More can improve quality but take longer. |
detailsEnable | boolean | No | true | - | - | - | Sharpens and enhances fine details in the result. On by default. |
baseModel | string | No | FLUX.1-dev | - | - | FLUX.1-dev, FLUX.1-Krea-dev | The base model to use: Stylized (Flux Dev) for artistic or illustrated images, Realism (Flux Krea Dev) for photographic ones. |
seed | number | No | - | 0 | 2147483647 | - | A number that makes results repeatable. Reusing the same seed and settings produces the same image; leave it empty for a random result each time. |