What is FLUX.1 Kontext? Features, pricing & FLUX.2
Quick answer
FLUX.1 Kontext is Black Forest Labs' previous-generation family for text-to-image generation and instruction-based editing. It still works through Pro, Max and downloadable Dev routes, but BFL recommends FLUX.2 for new projects. Use Kontext when an existing workflow depends on it; start a new evaluation with FLUX.2 unless a matched test shows a reason to stay.
Status, prices and license boundaries checked September 13, 2026 against the official Kontext overview, BFL pricing table and official repository.
FLUX.1 Kontext variants and current replacement
Variant | Current role | Official BFL API price | Commercial boundary |
|---|---|---|---|
Previous-generation hosted generation and editing | $0.04/image | BFL API terms | |
Previous-generation premium hosted route | $0.08/image | BFL API terms | |
Downloadable editing weights | Local compute; no hosted BFL API | Non-commercial weights; commercial licensing available | |
BFL's recommended generation and editing family for new projects | From $0.014-$0.07 by variant and resolution | Check the selected endpoint or weight license |
Prices are BFL's listed base API prices. Third-party hosts can charge different rates. Local Dev deployments still incur compute, storage and engineering cost.
Pro, Max or Dev?
Choose Pro for the lower listed hosted price and fast production iteration. Evaluate Max when the premium route's typography or adherence is worth twice the per-image price on your actual brief. Choose Dev when downloadable weights and customization matter enough to operate the model and the non-commercial license fits; contact BFL for commercial licensing rather than inferring permission from open weights.
When to move to FLUX.2
BFL now recommends FLUX.2 for new generation and editing work. Its documentation highlights multi-reference input, improved text editing and output up to 4 megapixels. Move when those capabilities matter, when starting a new integration, or when a current FLUX.2 variant beats Kontext on accepted-output cost.
Keep Kontext when compatibility, a fine-tuned workflow or a known production baseline has more value than migrating. Run the same source images and acceptance rubric before switching.
Existing visual examples from the original article
These six examples are retained from the original article as visual context. Their complete endpoints, seeds, candidate sets and untouched source files were not recorded, so they cannot prove a current quality, speed or consistency ranking.
Remove the mug from the right side of the desk. Reconstruct the revealed surface naturally while preserving the laptop, lighting, shadows, camera position, and every other object.
Try an instruction-based image edit
Upload one image, describe one controlled change, and inspect what stayed consistent before applying more edits.
Open AI Image EditorFor an editing workflow, prompt examples and a repeatable QA test, use our FLUX.1 Kontext tutorial.
Browser and model alternatives
Use Magic Hour AI Image Editor when you want a managed browser workflow and model choice rather than operating weights. The selected model and credit estimate can change, so verify the live option before assuming it is FLUX.1 Kontext. Our AI image editor comparison covers other editing workflows.
Frequently asked questions
It is still available, but Black Forest Labs labels it previous generation and recommends FLUX.2 for new projects.
BFL lists Pro at $0.04 per image and Max at $0.08 per image. Dev is local-only through BFL's current pricing table, so budget compute and operations separately.
BFL publishes downloadable open weights, but the official documentation states that Dev uses a non-commercial license and that commercial licensing is available. Open weights do not mean unrestricted commercial use.
It is designed for targeted instruction-based edits, but no prompt guarantees that all other pixels or details remain unchanged. Compare the entire output with the source.
Start new work with FLUX.2 because that is BFL's current recommendation. Retain Kontext when a matched evaluation, compatibility need or existing deployment justifies it.
