Sign up

Model schema

View how workloads are defined across models. Generate ready-to-run API requests with provider-aware parameters and execution controls.

bfl/flux-1.1-pro

Supported providers:
ReplicateFal
image11 fields

SCHEMA

Core generation parameters. These fields are normalized across all providers, we handle the mapping so you can use a single unified schema regardless of which provider runs your generation.

generation_promptstringrequired

Text description of the content you want to generate

generation_ratioenum

Aspect ratio of the generated content

default: "1:1"enum: 1:1 | 3:4 | 4:3 | 9:16 | 16:9
generation_output_formatenum

Format of the output file

default: "png"enum: png | jpg | webp
generation_output_numbernumber

Number of outputs to generate

min: 1max: 1default: 1
generation_provider_orderenum

Provider failover order

default: "replicate, fal"enum: replicate, fal | fal, replicate

SPECIFIC SCHEMA

Provider-specific parameters, only passed to providers that support them. Safe to omit; sensible defaults are applied automatically. For content moderation fields, we always default to the most permissive option to minimize blocked generations. Override only when you have a specific need.

generation_enhance_promptenum

Prompt enhancement mode - 'standard' for higher quality (may increase generation time), 'fast' for quicker enhancement, 'off' to disable

default: "off"enum: off | standard
generation_guidance_scalenumber

Classifier-free guidance scale - controls how closely the output follows your prompt (higher values increase prompt adherence)

min: 0max: 20default: 3.5
generation_moderationboolean

Content moderation (true = moderation on, false = moderation off)

default: false
generation_num_inference_stepsnumber

Number of denoising steps (more steps produce higher quality but slower results)

min: 1max: 100default: 28
generation_output_qualitynumber

Quality when saving output, higher levels mean best quality (not relevant for PNG)

min: 0max: 100default: 80
generation_seednumber

Random seed for reproducible generation (the same seed and prompt produce the same output every time), leave empty or use available default value for random seed