Runflow
Back to Gallery

black-forest-labs/flux-pro/kontext

FLUX.1 Kontext [pro] handles both text and reference images as inputs, seamlessly enabling targeted, local edits and complex transformations of entire scenes.

$0.04/image 2026-04-06

Input

Input 1

Example input image

Your request will cost$0.040

~25 runs per $1

Output

Example
A sleeping orange tabby cat added between the monitors on the desk

Example output from FLUX Kontext

Overview

FLUX Kontext [pro], 12B-param in-context editor. Character consistency, style transfer, local edits. $0.04/image.

Key capabilities

  • Character consistency without fine-tuning, Preserves subject identity, style, and brand elements across edits, no LoRA training or per-subject fine-tuning needed.
  • Up to 8× faster than competitors, Significantly faster inference than competing in-context editing models while maintaining high-quality results.
  • Local and global edits in one model, Handles everything from targeted region edits (change a shirt colour) to full scene transformations (move from studio to beach), no separate models or masks needed.
  • Multi-turn editing, Apply successive refinements that maintain visual coherence, each edit builds on the previous one without degrading quality.

FLUX Kontext model family by Black Forest Labs

ModelDetails
FLUX Kontext [max]Premium, best prompt adherence and typography, $0.08/image
FLUX Kontext [pro] (this model)Standard, fast in-context editing with character consistency, $0.04/image

Technical specifications

SpecDetails
ArchitectureRectified Flow Transformer with reference conditioning
Parameters12 billion
ModeImage-to-image (in-context editing)
SpeedUp to 8× faster than state-of-the-art models
Guidance Scale3.5 (default) · 1–20 range
Max Batch Size4 images per request
Input FormatsJPG · JPEG · PNG · WebP · GIF · AVIF
Output FormatsJPEG · PNG
LicenseCommercial

Frequently asked questions

Related models

Start generating with FLUX Kontext

Get API access in minutes. No GPU setup, no infrastructure to manage.