Andrew AdamsAndrew Adams · Co-Founder & Operations at Wireflow ·

FLUX 3: What It Does and How to Run It in Wireflow

Black Forest Labs announced FLUX 3 on 23 July 2026.

Here is what is live, what is still gated, and how to build the flow in Wireflow today.

Free to build · no credit card · See how it works

Flux Text to Image (Flux 2 Pro, Flux 3 ready)Open workflow →
FLUX 3: What It Does and How to Run It in Wireflow
Loading interactive canvas…
750+Built on 750+ internal test generations during development
10+10+ AI models benchmarked for optimal output quality
30+30+ configurations tested to find the best defaults

We've run 750+ flux 3: what it does and how to run it in wireflow generations internally while building Wireflow and identified the three factors that separate high-quality AI outputs from generic ones — and built them directly into this workflow.

01How it works

How to Use FLUX 3: What It Does and How to Run It in Wireflow

Steps to get you started in Wireflow.

Add the Nodes
Step 1

Add the Nodes

Open the Wireflow canvas, drop in a text input node for your prompt, and add an image model node next to it. Nothing else is needed to start.

Wire It Up and Run
Step 2

Wire It Up and Run

Connect the prompt output to the model prompt input, pick FLUX 2 Pro, and run the flow. The generated image appears on the model node.

Swap the Model Later
Step 3

Swap the Model Later

When FLUX 3 Image is released and added to the model list, change the model on the same node. Your prompt, wiring and downstream nodes stay as they are.

02

What FLUX 3 Is

FLUX 3 is Black Forest Labs' third generation model family, announced on 23 July 2026. Unlike FLUX 2, which is an image model, FLUX 3 is described as a single multimodal flow model that spans image, video, audio and robot action prediction. Black Forest Labs calls this direction Real World Models. The family has four named tiers: FLUX 3 Video, FLUX 3 Image, FLUX-mimic for robotics, and an open weight FLUX 3 Dev.

Only the video tier is reachable today, through gated early access on the Black Forest Labs site. FLUX 3 Image has been announced but not released, there is no published pricing, and no third party host carries a FLUX 3 endpoint yet. That means the honest answer for anyone building right now is to wire the flow with a model you can already run. In Wireflow you build a prompt node into an AI image generator node, run it on FLUX 2 Pro, and swap the model when FLUX 3 arrives. The rest of the graph, including any downstream nodes you add through the orchestration API, does not change.

03

What Black Forest Labs Announced

01

Video Tier in Early Access

FLUX 3 Video covers text to video, image to video, video to video and keyframe to video, behind a gated signup.

02

Clips Up to 20 Seconds

The video tier generates clips up to 20 seconds long, with preliminary evaluations run at 720p.

03

Native Synchronized Audio

Audio is generated inside the same model rather than added afterward, including multilingual dialogue.

04

Agentic Clip Chaining

Black Forest Labs describes chaining clips together agentically to build longer sequences from one model.

05

Image Tier Announced Only

FLUX 3 Image promises complex prompt handling, multilingual text rendering and editing, but is not released.

06

FLUX-mimic for Robotics

A separate tier applies the same flow model to robot action prediction rather than to media generation.

More Than Just FLUX 3: What It Does and How to Run It in Wireflow

Run FLUX Today, Not Someday

FLUX 3 Image is not out yet, so build on what ships now. The same canvas already runs FLUX 2 Pro.

Run FLUX Today, Not Someday

Swap Models, Keep the Wiring

Changing the model on a node leaves prompts and connections intact, just like a switch to Nano Banana.

Swap Models, Keep the Wiring

Call the Canvas From Code

Every published flow runs over REST. Post to execute, then poll for the image URL via the workflow API.

Call the Canvas From Code

Chain Image Into Video

Feed an image node into a video node on the same canvas, the pattern behind our AI video generator pages.

Chain Image Into Video

Honest Release Tracking

We label what is live and what is still gated, the same as we do for FLUX Pro, so you never wait on a dead node.

Honest Release Tracking
15+

AI Models Available

API Access

Automate Any Workflow

Monthly Credits

Included in Every Plan

FAQs

FLUX 3 is Black Forest Labs' multimodal flow model announced on 23 July 2026. One model covers image, video, audio and robot action prediction, released in separate tiers rather than as a single product.

Andrew Adams

Written by

Andrew Adams · Co-Founder & Operations at Wireflow

Runs client operations and content strategy at Wireflow. Works directly with creative teams and agencies to build production AI workflows.

Content StrategyClient Operations

Build the FLUX Flow Now

Open the prompt to image flow on the Wireflow canvas. It runs on FLUX 2 Pro today, and the model on that node is the only thing you change when FLUX 3 opens up.

Free to buildNo credit cardNo GPU or installCancel anytime