Can FLUX 3 generate from text only?

Yes. A prompt is always required, while images and video are optional input strategies.

How do image inputs work?

One image defines the opening frame, two define start and end, and three to ten form an ordered storyboard. Choose an explicit duration for storyboard generation.

Can I combine images with a starting video?

No. Video continuation accepts one MP4 up to 50MB and 15 seconds and cannot be combined with image inputs.

It is a lower-cost preview for text-to-video or image-to-video. It is not available for video continuation.

Is FLUX 3 production-ready?

FLUX 3 is currently an early-access preview. Treat outputs as candidates that still require visual, audio, identity, and rights review.

Black Forest Labs early access

Turn a prompt, storyboard, or clip into a FLUX 3 video

Create a new scene from text, control a sequence with up to ten ordered images, or continue an existing video. Choose 5–20 seconds, 720p or 1080p, synchronized audio, and draft 720p for cheaper iteration.

Open FLUX 3

Open video generation with this model selected

See examples and review points

Video
Black Forest Labs
5s-20s
Storyboard

Visual examples

Compare real FLUX 3 motion across four different briefs

Play the official example alongside three 5-second Draft 720p outputs generated through the Toono dashboard. Compare product realism, camera control, scene continuity, and native sound before spending on a longer or higher-resolution run.

Black Forest Labs

Official FLUX 3 example

Review the futuristic San Francisco clip for movement, continuity, camera behavior, and environmental audio.

Black Forest Labs

Beverage launch shot

An amber bottle tests glass reflections, condensation, a controlled orbit, and subtle studio sound.

Black Forest Labs

Outdoor product motion

A rain-soaked trail shoe tests material detail, restrained camera movement, and object stability.

Black Forest Labs

Near-future worldbuilding

A desert weather station tests environmental motion, stable geometry, atmosphere, and scale.

The three Toono examples below were generated from text with Draft 720p, 16:9, and audio enabled. Image storyboards and video continuation remain available as separate input modes in the workspace.

Practical workflow

Choose one input strategy, then direct the sequence

FLUX 3 accepts text, ordered images, or a starting video. A clear input role and a concise action arc make the result easier to evaluate.

01
Choose text, images, or video
Start from a prompt, an ordered 1–10 image sequence, or one MP4 continuation source.
02
Set timing and framing
Choose 5–20 seconds, the intended aspect ratio, resolution, and whether generated audio is useful.
03
Draft, review, and render
Validate the idea cheaply in draft 720p, then rerun the strongest direction at full quality.