Black Forest Labs has spent two generations of FLUX becoming one of the best-known names in AI image generation. Flux 3 is the first time they have pointed that technology at video. It also does something many video models still treat as a second step: it creates sound in the same pass as the picture. Flux 3 is available on Apiframe now, behind the same endpoint and API key as the rest of the catalog.
What Is Flux 3?
FLUX 1 and FLUX 2 were image models, and they are still the right choice when you only need a still image. Flux 3 is a different kind of model. It is one model trained on images, video, and audio together, instead of three separate systems joined into one pipeline. Black Forest Labs announced it on July 23, 2026. It builds on Self-Flow, the company’s method for teaching one model to both create and understand different kinds of media.
That design has a practical benefit. When audio and video come out of the same pass, they line up. An impact sound lands on the exact frame where two objects meet, and lip movement matches the spoken words. Models that make a silent clip and add sound afterward have to guess the timing later, and that is usually where you notice a mismatch. For a closer look at how other models handle this, see our comparison of APIs that support synchronized audio with generated video.
The same model also powers FLUX-mimic, a robot control model that Black Forest Labs built with mimic robotics. Audi is testing it in its production lab. It hints at how much the model had to learn about touch, weight, and cause and effect just to predict video accurately.
Pricing That Scales With the Clip
Like most video models, Flux 3 bills per second of output: 25 credits per second at 720p and 42 credits per second at 1080p, with audio included in the rate. Ask for a 5-second clip and you pay for 5 seconds. Ask for 20 and you pay for 20.
One thing to know: duration is optional. Leave it out and the model picks a length that suits the content. Credits are deducted when you submit the job, before that length is known, so an omitted duration is billed at the 20-second maximum. Pass an explicit duration whenever you want exact, predictable pricing.
Total cost is simply the rate multiplied by the clip length, so a 10-second 1080p clip costs 420 credits. Continuation costs more than generating from scratch, since the model has to read your source video and stay faithful to it while generating new footage. The full rate table is further down this page, and you only pay for successful generations. Our model documentation has the complete parameter reference and current rates. For how this compares with the rest of the market, see our breakdown of AI video API pricing in 2026, and our look at which APIs support synchronized audio with generated video.
What You Can Actually Do With It
Text to video, with sound
A prompt on its own gives you a clip with sound already in place: background noise, footsteps, speech, music, whatever the scene suggests. Audio is on by default. Set generate_audio to false if you plan to add your own sound later.
Animate a single image
Send one image in images and it becomes the first frame. This is the way to add motion to artwork you have already approved, such as a product photo, a character render, or a Flux 2 image. You keep the look and generate only the movement. Our image-to-video API tutorial shows this in code, and our Flux API guide covers the earlier image models you can use to make those stills.
First and last frame
Send two images and Flux 3 treats them as the opening and closing frames, then creates the movement in between. This is the most controllable mode. You decide where the shot starts and ends, and the model fills in the middle.
Storyboard from up to ten images
Three or more images work as an ordered storyboard, and the model paces the clip to hit each image in order. A storyboard has no built-in length, so you must set duration yourself. Give each moment enough time to land.
Continue an existing clip
Point start_video at an MP4 file (up to 50 MB and 15 seconds), and Flux 3 continues from its final frames. It carries the scene, characters, and audio forward instead of cutting to something new. You can chain this by feeding each result back in as the next start_video, which lets you build sequences well past the 20-second limit of a single generation. You cannot use start_video and images in the same request. Continuation costs more than a fresh clip, as the pricing section below explains.
Draft mode
Set draft to true for a quick, low-cost preview at about a third of the price. Drafts are always 720p, so the resolution setting is ignored. Use drafts to find the right composition and camera move, then run your best prompt again at full quality with one setting changed.
Specs at a Glance
| Spec | Detail |
|---|---|
| Output duration | 5 to 20 seconds, any whole second. Omit it and the model picks a length to fit the content (billed at the 20-second maximum) |
| Resolution | 720p or 1080p. Drafts are always 720p |
| Audio | Synchronized audio, generated in the same pass, on by default |
| Aspect ratios | 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16, or auto |
| Image inputs | Up to 10: one for a first frame, two for opening and closing frames, three or more for a storyboard |
| Video input | One MP4 up to 50 MB and 15 seconds, for continuation |
| Prompt length | Up to 4,000 characters |
| Pricing | Per second of output; the rate depends on resolution, draft mode, and continuation |
Where This Actually Gets Used
- Paid social and ads: A 20-second cut that already has audio is ready to publish, not a silent clip waiting for a sound pass.
- Product motion from existing photos: Animate photography that has already been approved, instead of regenerating the product and risking a version that looks different.
- Multi-shot stories: Chain continuations to run a scene past a single generation, with characters and audio carrying across the cuts.
- Storyboard previews: Hand over an ordered set of frames and get back a timed rough cut that hits each beat.
- Trying many ideas cheaply: Draft mode makes it affordable to test 50 directions before you commit to a full-quality render.
How to Use Flux 3 on Apiframe
Flux 3 uses the same endpoint as every other video model on Apiframe. If you're already generating video here, this is a one-line change:
curl -X POST https://api.apiframe.ai/v2/videos/generate \
-H "X-API-Key: afk_your_api_key_here" \
-H "Content-Type: application/json" \
-d '{
"prompt": "a lighthouse keeper climbs a spiral staircase at dawn, lantern swinging, waves booming against the rocks outside",
"model": "flux-3",
"flux3Params": {
"duration": 20,
"resolution": "1080p",
"aspect_ratio": "16:9"
}
}'You get a job ID back immediately:
{
"jobId": "c3d4e5f6-a7b8-9012-cdef-345678901234",
"status": "QUEUED"
}Poll /v2/jobs/:id, or set a webhookUrl on the original request and let Apiframe call you when the clip is ready:
{
"id": "c3d4e5f6-a7b8-9012-cdef-345678901234",
"status": "COMPLETED",
"model": "flux-3",
"result": {
"videoUrl": "https://cdn2.apiframe.ai/videos/c3d4e5f6-a7b8-9012-cdef-345678901234.mp4"
}
}No Black Forest Labs early-access application, no second SDK, no new billing relationship. Flux 3 sits in the same catalog, credit balance, and job API as everything else.
Pricing
| Generation | Credits per second | |
|---|---|---|
| 720p | 29 | |
| 1080p | 49 | |
| Draft | 10 | |
| Continuation, 720p | 59 | |
| Continuation, 1080p | 90 | |
| Continuation, draft | 17 |
Total cost is the rate times the clip length, so a 10-second 1080p clip is 420 credits. Omitting the duration bills at the 20-second maximum, so set it explicitly for exact pricing. Continuation costs more because the model has to read your source video and stay faithful to it while generating new footage. As always, you only pay for successful generations. The model docs page has the full parameter reference, and our guide to estimating the monthly cost of AI video generation turns these rates into a budget you can plan against.
Try It Today
Flux 3 is available now on Apiframe, alongside every other major video model, behind one endpoint and one API key. Get an API key and generate your first clip, or read the docs for full details on every setting.
Want to see how it stacks up? Read our Flux 3 vs Hailuo 3 comparison, look at Hailuo 03 on Apiframe, or browse our roundup of the best AI video generation APIs. New to making video from code? Start with our AI video generation API guide. Questions? Join our Discord.
Apiframe Team
The team behind Apiframe - making AI generation accessible to everyone.