Flux AI logoFlux AI
FLUX 3 early access

FLUX 3: The Multimodal AI Model from Black Forest Labs

Explore the announced FLUX 3 capabilities for images, video, and audio, then turn your next visual idea into an image in Flux AI.

What is FLUX 3?

FLUX 3 is Black Forest Labs' announced multimodal foundation model. It is designed to learn from images, video, and audio in one architecture and is currently rolling out through early access.

One multimodal foundation

FLUX 3 is built around a shared representation for visual and audio information, rather than separate models for each medium.

Image, video, and audio direction

The official release describes text-led generation plus image and video references, with video-and-audio output among its planned capabilities.

Image creation and editing

Black Forest Labs says FLUX 3 can synthesize and edit images across styles, aspect ratios, and resolutions, with early access still in progress.

Why FLUX 3 matters for visual creation

Work from text or visual references

Work from text or visual references

The FLUX 3 announcement highlights generation from text prompts as well as input references such as images and video - useful when an idea needs a clear visual starting point.

Complex prompts and text in images

Complex prompts and text in images

According to Black Forest Labs' preliminary evaluation, FLUX 3 improves on earlier FLUX generations for complex prompts and text rendering in images.

A platform direction, not a static tool

A platform direction, not a static tool

The release frames FLUX 3 as a foundation for creative generation and editing across modalities. Availability, features, and results may change during early access.

How to start creating while FLUX 3 rolls out

FLUX 3 access is being introduced by Black Forest Labs. You can use the same clear prompt workflow for image ideas in the Flux AI canvas today.

  1. 1

    Describe the outcome

    State the subject, setting, composition, lighting, and visual style. For a reference-driven concept, explain what should remain consistent.

  2. 2

    Create an image in Flux AI

    Use the canvas above to generate an image with the available Flux AI models. FLUX 3 itself is not yet a selectable model on this site.

  3. 3

    Refine and download

    Adjust the prompt for the next variation, then download the result when it matches your creative direction.

How to start creating while FLUX 3 rolls out

Explore related AI image workflows

Use the current Flux AI tools for the image-generation tasks that align with the FLUX 3 search intent.

Text-to-image concepts

Turn a written brief into original images for mood boards, campaigns, and early creative exploration.

Reference-based image edits

Start from an uploaded image and direct a new look, scene, or visual treatment with a precise edit prompt.

Prompt development

Build a clearer visual prompt before generating, so the subject, composition, and style are easier to control.

Flux AI Models

Choose from multiple powerful variants AI Image Model of Flux AI

Ghibli Image Generator

Ghibli Image Generator

Create Studio Ghibli style images with our AI Ghibli generator. Transform your ideas into Ghibli style art in seconds.

Try model
Flux Pro

Flux Pro

Flux1.1 [pro] is the flagship Flux AI generator model, optimized for high-end commercial use.

Try model
Flux 1.1 Pro Ultra

Flux 1.1 Pro Ultra

FLUX 1.1 Pro Ultra is 4k resolution ai image generator, optimized for high-speed generation.

Try model
Flux Schnell

Flux Schnell

Flux.1 [Schnell] is turbo mode of Flux AI generator, the fastest model in the Flux AI

Try model
Flux Dev

Flux Dev

FLUX.1 [dev] is a 12 billion parameter rectified flow transformer that can generate images from text descriptions.

Try model
Flux Image Generator

Flux Image Generator

Create high-quality AI images from a text prompt.

Try model
Stable Diffusion 3.5 large

Stable Diffusion 3.5 large

SD3.5 large is superior quality and prompt adherence, this base model is the most powerful in the Stable Diffusion family.

Try model
IC Light v2

IC Light v2

IC Light V2 is an advanced image processing tool that utilizes a 16-channel Variational Autoencoder (VAE) to enhance and manipulate lighting in images.

Try model

FLUX 3 FAQ

Key details from the official FLUX 3 announcement and how this page relates to Flux AI.

Turn your FLUX 3 ideas into an image

Start with the Flux AI image tools available today while you follow FLUX 3's official early-access rollout.

Create an image