Flux.3 VideoAI video generator

Native audio, up to 20 s. Pay only for the takes you keep — kill the rest for a full refund.

Text to videoImage to videoNative audio
Only on Elyum · Keep what you love. Unique keep/kill system with a full refund on kills.
Try Flux.3 Video freeSee pricingOur algorithm may grant you 100 free credits · no card required
Real Flux.3 Video output, rendered on Elyum — untouched. Prompt: “A glacier face calves into arctic water, ice crystals sparkling in the air, deep blue and white tones, slow majestic scale.”
Developed by Black Forest Labs
Max resolution1080P
Clip length5–20 seconds
Aspect ratios7 options

Flux.3 Video is Black Forest Labs' video model: 5–20 second clips with native sound and dialogue, from text or a still. It is the video release of FLUX 3, the lab's first video model, announced July 23, 2026 and generally available from August 4, 2026. On Elyum it does text-to-video and image-to-video.

Use it when you need a talking clip with the line already in the file, a still that should move, or a short that switches camera inside one take. Web studio, MCP, or REST — one credit balance.

Flux.3 Video pricing on Elyum

ResolutionCredits / second5s clip15s clip
720p23115 credits · $1.15345 credits · $3.45
1080p34170 credits · $1.70510 credits · $5.10

1 credit = $0.01. The struck-through number is the pre-cut rate — the discount is already in the live price. You only pay for takes you keep; previews and killed takes cost nothing. Full plan details on the pricing page.

Why Flux.3 Video is good

Native audio in the same take

Dialogue, effects, and ambience generate with the frames. Quote the spoken line in the prompt if you want speech. You do not stitch a silent clip and a separate soundtrack.

5–20 seconds at 720p or 1080p

Clips run 5 to 20 seconds in one generation at 720p or 1080p. Seven aspect ratios. Rate is per kept second: 23 credits/s at 720p, 34 at 1080p.

Text to video and image to video

Describe the scene in plain language or a detailed prompt, or pin a still as the opening frame. On Elyum those are the two routes. Native audio is on both.

Multilingual speech, lip-synced

BFL lists lip-synced speech in English (several dialects), Chinese, Spanish, French, German, Japanese, Portuguese, Russian, Italian, Indonesian, Turkish, Hindi, Punjabi, and more. Put the line in the prompt.

Multiple shots in one take

The model can change scene and camera angle inside a single clip and keep the sequence coherent. BFL also notes on-screen type as part of the picture, not a later overlay.

Pay only for keeps

Preview is free. A 5-second 720p keep is 115 credits ($1.15); 15 seconds is 345 credits ($3.45). Kill refunds the rest.

Why generate with Elyum?

  • Pay only for what you keep — full refund on kills.
  • Every top video model in one place, one balance.
  • Rates ~20% under comparable platforms.
  • Kept files download clean and full quality — no watermark, commercial use included.
  • Works from the studio, or from Claude, ChatGPT and Cursor via MCP.
  • Free plan — our algorithm may grant you 100 credits, no card required.

How it works

  1. Sign up free on Elyum. The Free plan includes up to 100 credits (our algorithm may grant them); every preview is free until you keep.
  2. Open the video studio and pick Flux.3 Video.
  3. Write the prompt, or attach a still for image-to-video. Set length between 5 and 20 seconds and pick 720p or 1080p.
  4. Run the take. Watch the preview with native audio.
  5. Keep the ones you want — credits charge then — or kill the rest for a full refund. Downloads are clean, no watermark, commercial use included.

What people make with it

  • Presenter or character dialogue clips with lip-sync already in the file
  • Image-to-video from a product, character, or location still
  • Social and ad cuts that switch camera inside a 5–20 second take
  • Multilingual spots without a separate voice-over pass
  • Stylized, cinematic, or motion-design B-roll, not only photoreal
  • Short educational clips from a written brief

Frequently asked questions

Is Flux.3 Video free to try?
Yes. Every render comes back as a free preview. The Free plan includes up to 100 credits (our algorithm may grant them). You only pay when you keep a take; kill refunds the rest in full.
How much does Flux.3 Video cost on Elyum?
You pay per kept second, not per prompt. 720p is 23 credits per second — 115 credits ($1.15) for 5 seconds, 345 credits ($3.45) for 15. 1080p is 34 credits per second — 170 credits ($1.70) / 510 credits ($5.10). Killed takes refund in full.
Does Flux.3 Video generate audio with the video?
Yes. Sound and dialogue are native to the clip: speech, effects, and ambience in the same pass as the picture. Prompt the spoken line if you want speech. You preview the audio with the take, then keep or kill as usual.
What can Flux.3 Video do besides text to video?
On Elyum it does text to video and image to video. Clips run 5–20 seconds at 720p or 1080p, with seven aspect ratios. Native audio is on both.
Can I use Flux.3 Video videos commercially?
Kept takes download at full quality with no watermark. Commercial use is included on Elyum. Preview files are for judging, not for shipping.
Flux.3 Video vs other AI video models on Elyum?
Flux.3 Video is the Black Forest Labs pick: native audio, 5–20 second clips from text or a still, including multilingual speech. Other studio models trade that for longer clips, more references, or a different motion grammar. Same credit balance either way.

More models to try

Create everything. Keep the best.

Flux.3 Video and 50+ other top models on one balance. Every render is a free preview — pay nothing for the rest.

Start creating free