FLUX.1 [dev] Ultra-Fast
FLUX.1 [dev] Ultra-Fast is an image generation model from Black Forest Labs optimized for speed with LoRA and inpainting support.
Black Forest Labs
WaveSpeedFast FLUX image generation with LoRA support
FLUX.1 [dev] Ultra-Fast is an image generation model developed by Black Forest Labs and served through WaveSpeed AI's infrastructure, designed to produce images at reduced latency compared to standard FLUX.1 [dev] deployments. It supports LoRA (Low-Rank Adaptation) fine-tuning, allowing users to apply custom style or subject adapters at inference time, and accepts a source image input alongside a mask for inpainting workflows. The model is priced at $0.006 per image and is available on MindStudio without requiring separate API credentials.
This variant is well suited for workflows that require iterative image generation, style-consistent outputs via LoRA adapters, or targeted image editing through masked inpainting. Users can control output dimensions, inference steps, guidance scale, and denoising strength, giving fine-grained control over the generation process. The 10,000-token context window applies to the text prompt input, supporting detailed scene descriptions. It is a practical choice for developers building AI applications that need image generation with customization options at a predictable per-image cost.
What FLUX.1 [dev] Ultra-Fast supports
LoRA Adapter Support
Accepts one or more LoRA adapters at inference time, enabling custom styles or subjects without retraining the base model.
Inpainting
Takes a source image and a mask image as inputs to selectively regenerate only the masked region of an existing image.
Image-to-Image
Accepts a source image URL and a strength parameter to guide how much the output deviates from the input image.
Configurable Output Size
Exposes separate width and height inputs so developers can specify exact pixel dimensions for generated images.
Inference Step Control
Allows setting the number of diffusion steps and guidance scale, giving control over quality and prompt adherence per generation.
Reproducible Generation
Supports a seed input so that identical prompts and settings produce consistent, reproducible image outputs.
Long Prompt Context
Supports up to 10,000 tokens in the text prompt, accommodating detailed scene descriptions and complex generation instructions.
Ready to build with FLUX.1 [dev] Ultra-Fast?
Get Started FreeCommon questions about FLUX.1 [dev] Ultra-Fast
How much does FLUX.1 [dev] Ultra-Fast cost per image?
The model is priced at $0.006 per image generated.
What is the context window for this model?
The model supports a context window of 10,000 tokens, which applies to the text prompt describing the image to generate.
Does this model support LoRA adapters?
Yes. The model includes a dedicated LoRAs input that accepts one or more LoRA adapters, allowing custom styles or subjects to be applied at inference time without modifying the base model.
Can I use this model for inpainting?
Yes. The model accepts both a source image and a mask image as inputs, enabling inpainting workflows where only the masked area of an existing image is regenerated.
Who publishes this model and who provides the inference?
The model is published by Black Forest Labs and served through WaveSpeed AI's inference infrastructure, which is the source of the ultra-fast generation speed.
Is there a knowledge cutoff date for this model?
FLUX.1 [dev] Ultra-Fast is an image generation model and does not have a knowledge cutoff in the traditional sense. It generates images from prompts rather than answering questions based on training data.
Documentation & links
Parameters & options
Optional reference image for image-to-image generation.
The mask image tells the model where to generate new pixels (white) and where to preserve the original image (black). It acts as a stencil for targeted image editing.
The number of denoising steps. More steps generally produce higher quality but take longer.
Controls how closely the generation follows the prompt. Higher values produce results closer to the prompt but may reduce diversity.
Controls the extent to transform the reference image. Only applies when an image is provided. 1.0 means full transformation.
Explore similar models
Black Forest LabsFLUX.2 [klein] 9B
FLUX.2 [klein] 9B is a 9-billion-parameter image generation model from Black Forest Labs released in January 2026.
Black Forest LabsFLUX.1 [schnell] LoRA
FLUX.1 [schnell] LoRA is an image generation model from Black Forest Labs that supports custom LoRA adapters for stylized output.
Black Forest LabsFLUX.1 [dev] LoRA
FLUX.1 [dev] LoRA is an image generation model from Black Forest Labs that supports custom LoRA adapters and inpainting.
QwenQwen Image
Qwen Image is an image generation model from Qwen that supports LoRA adapters and source image inputs for guided generation.
X.aiGrok Imagine 2.0
Grok Imagine 2.0 is an image generation model from X.ai that accepts source images and supports configurable aspect ratio, resolution, and quality.
ByteDanceSeedream 5.0 Pro
Seedream 5.0 Pro is an image generation model from ByteDance that supports source image input and flexible aspect ratio and resolution controls.
Start building with FLUX.1 [dev] Ultra-Fast
No API keys required. Create AI-powered workflows with FLUX.1 [dev] Ultra-Fast in minutes — free.