FLUX 3 Max
FLUX 3 Max is a post-trained variant of Black Forest Labs' FLUX 3 Video, adapted and served by fal for stronger prompt adherence and visual quality at much lower latency. It was announced on 19 October 2026 with three fal endpoints: text-to-video, image-to-video with an optional last frame, and keyframes-to-video. Clips run 5 to 20 seconds at 720p or 1080p with native audio on by default, and a reasoning setting lets the model expand a short prompt before it renders. BFL says it is tied for first place in image-to-video and makes a 10-second clip in about 12 seconds.
Anvisha Pai, Co-founder & CEO, Voyager
fal / Black Forest Labs · released
›What we asked for
A self-portrait of FLUX 3 Max: a single continuous shot racing through a pine forest at dawn as light streaks between the trunks, mist lifting, birdsong and wind, cinematic and realistic.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
At a glance
- Price
- $0.17 a second at 720p and $0.29 at 1080p on fal, audio included, read October 19, 2026. A 75% launch promotion, $0.0425 at 720p, runs through November 7.
- Measured wait
- About 16 seconds from queue submission to a five-second 720p clip on eight of nine prompts, 39 seconds on the ninth. One batch, not a service guarantee.
- Length
- 5 to 20 seconds, at 720p or 1080p.
- Inputs
- Text, a first frame with an optional last frame, or a list of keyframes placed at chosen frames.
- Audio
- Native audio, on by default.
- Ranking
- Black Forest Labs says it is tied for first in image-to-video and 22 times faster than Seedance 2.5 at the same quality. That is the vendor's own evaluation.
Other FLUX models: FLUX 3 Image, FLUX.2 Pro, FLUX.2 Flex, FLUX.2 Max. Compare all 5.
Anvisha's take
FLUX 3 Max is the most surprising run on the roster: every one of the nine clips came back in about 16 seconds, where Seedance 2.5 took 3 to 26 minutes on the same prompts. Speed did not cost the look. The spill is the most convincing physics clip I have from any model, with the water pooling and the cubes sliding apart; the market is dense and legible, with the dog and the busker both present; the title card spells both lines; and the woodblock lighthouse stays a print as the wave breaks. Its misses are familiar ones: the turntable dripper grows a handle, and in image-to-video it kept our product exactly but filled it like a cup. At $0.17 a second at 720p it costs about a third of Seedance 2.5, and a quarter of that until November 7. For cinematic and realistic shots I would start here. These are frame-based observations; audio was generated on every clip but not scored by ear.
FLUX 3 Max real output: the same 9 clips every model gets
Every video model on this site gets the same 9 clips, so you can compare like with like. These are the clips FLUX 3 Max returned, one try each, nothing re-rolled or retouched. Each tile shows one frame of the clip until you press play. Generated on . The 9 clips took 16 to 39 seconds each and cost $1.91 in total.
A clean product shot that holds its shape through a full slow rotation.
›What we asked for
A matte black ceramic pour-over coffee dripper rotating slowly on a white turntable, soft studio light, seamless white background, product video, no text.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
A face, lip sync and a spoken line, where the model makes audio.
›What we asked for
A woman in her 40s with short grey hair stands in a bright kitchen, looks at the camera and says 'Good morning, the coffee is ready', natural window light, handheld feel, she smiles at the end.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
Set text that stays spelled correctly while it moves.
›What we asked for
A title card reading 'MODA SUMMER SESSIONS' animates in over slow-motion footage of a festival crowd at dusk, then a smaller line 'Friday 12 June' fades in beneath it. Bold condensed type, tangerine on navy.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
A steady drone push-in with stable geometry.
›What we asked for
Drone shot pushing slowly forward over a coastal town at sunrise, terracotta roofs, fishing boats in the harbour, long shadows, cinematic and steady.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
A spill in slow motion: does water behave like water.
›What we asked for
A glass of water tips over on a wooden table and spills in slow motion, water and three ice cubes sliding across the wood, macro lens, morning light.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
Hands, a kettle and a steady pour without extra fingers or teleporting objects.
›What we asked for
Close-up of hands pouring hot water from a gooseneck kettle in slow circles into a pour-over coffee dripper, steam rising, shallow depth of field, warm cafe light.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
Many people and objects moving at once without merging.
›What we asked for
A crowded Saturday farmers market seen from slightly above, shoppers with tote bags, a dog on a leash, a busker with a guitar, bunting swaying, late morning light, gentle camera drift.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
A named style held across every frame rather than drifting to photoreal.
›What we asked for
A lighthouse on a rocky cliff in a storm, in the style of a Japanese ukiyo-e woodblock print, waves crashing in flat colour with visible wood grain, animated like moving paper cutouts.
blackforestlabs/flux-3-max/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 704 pixels
Bring a still product photo to life without changing the product.
›What we asked for
The coffee dripper slowly fills as coffee drips through it, steam rising, the camera drifting gently around it, the dripper itself unchanged. (starting from The Nano Banana Pro product shot from the image fixture set.)
blackforestlabs/flux-3-max/image-to-video
aspect_ratio: auto, resolution: 720p, duration: 5, generate_audio: true
returned 960 by 960 pixels
How the cost under each picture was worked out: fal FLUX 3 Max 720p $0.0425/second with audio for 5 requested seconds; fal's 75% launch promotion, October 15 to November 7, 2026, regular $0.17 (fal.ai/models/blackforestlabs/flux-3-max/text-to-video, read 2026-10-19); fal does not return the charge.
Compare FLUX 3 Max
FLUX 3 Max pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | Second at 720p fal's regular rate, audio included. fal's 75% launch promotion, $0.0425 a second at 720p and $0.0725 at 1080p, runs October 15 to November 7, 2026, then these regular rates apply. | $0.17 | fal model page, read |
| fal | Second at 1080p | $0.29 | fal model page, read |
| fal | Five-second 720p clip Our clips cost $0.2125 each at the promotional rate. | $0.85 | fal model page, read |
| Voyager | Second at 720p 17 credits a second at 720p and 29 at 1080p, 1 credit for 1¢ of what fal charges, so the launch promotion passes through while it runs. On every paid plan; Free has no video. | $0.17 | Voyager pricing, read |
Is FLUX 3 Max free?
The fal endpoints are paid per second of video. We have not found a recurring free allowance for this model.
- •fal's launch promotion cuts the price by 75% through November 7, 2026; it is a discount, not a free tier.
fal model page, read
Where to use FLUX 3 Max
- fal, text-to-video · Hosted API and playground; the endpoint used for eight of the clips below.
- fal, image-to-video · First frame plus an optional last frame; used for the image-to-video clip.
- fal, keyframes-to-video · Several images pinned to chosen frames. Not tested here.
- Voyager · Runs inside Voyager, the agent for creative work, on Voyager credits on every paid plan, or with your own fal key.
Voyager
Run FLUX 3 Max inside Voyager
Voyager is an agent for creative work that runs every model in this reference, FLUX 3 Max included. Brief it, and it picks the model, makes the thing and hands back something you can edit. Create an account to get started.
What we noticed running FLUX 3 Max
- •The glass tips, the water pools across the table and the cubes slide apart: the most convincing spill in our set. See the physics clip.
- •The dripper gains an unrequested handle as it turns, as it did for H3 Max and Veo 3.1 Fast. See the product turntable clip.
- •The photographed dripper is kept exactly, but coffee pours in from above and fills it like a cup instead of dripping through. See the image to video clip.
- •Both lines of the title spelled correctly, set over a navy wash on the crowd. See the animated title card clip.
FLUX 3 Max limits and API parameters
| Limit | Value |
|---|---|
| Duration | 5 to 20 seconds, in whole seconds; 720p or 1080p. fal API schema, read |
| Output size | Our 16:9 clips came back at 1280 by 704, 24 frames a second, 5.04 seconds long; the square product photo gave a 960 by 960 clip. fal API schema, read |
›Every setting, for developers
fal (blackforestlabs/flux-3-max/text-to-video)
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Free text Required. | |
| duration | integer | 5 to 20 | 5 |
| resolution | enum | 720p, 1080p | 720p |
| aspect_ratio | enum | auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16 auto is 16:9 for text prompts and follows the first image otherwise. | auto |
| generate_audio | boolean | true, false | true |
| reasoning_level | enum | low, high How much the model expands your prompt before rendering; high is slower. | high |
| seed | integer | Any Random when not set. |
FLUX 3 Max image-to-video
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| image_url | URL | PNG, JPEG or WebP Required; the first frame. | |
| end_image_url | URL | PNG, JPEG or WebP Optional; the last frame. |
FLUX 3 Max keyframes-to-video
fal API schema, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| keyframes | array | image_url plus frame_index for each Required. Frame indexes run up to the clip's duration times its frame rate. |
What FLUX 3 Max will not make
The hosted endpoints apply fal's content checks. None of our nine prompts was refused, including the image-to-video product photo that Seedance 2.5 declined. Review the provider terms for your intended use.
fal model page, read
Alternatives to FLUX 3 Max
Frequently asked questions
How much does FLUX 3 Max cost?
On fal, $0.17 a second at 720p and $0.29 at 1080p, audio included, read October 19, 2026: $0.85 for a five-second 720p clip. fal's launch promotion takes 75% off through November 7, 2026, so our clips cost $0.2125 each. In Voyager it is the same rate in credits, 17 a second at 720p, on any paid plan.
How fast is FLUX 3 Max?
Eight of our nine five-second clips came back about 16 seconds after we submitted them, the ninth in 39. Black Forest Labs says a 10-second clip takes about 12 seconds. Seedance 2.5 took between 3 and 26 minutes on the same prompts.
What is the difference between FLUX 3 and FLUX 3 Max?
FLUX 3 Max is fal's post-trained version of Black Forest Labs' FLUX 3 Video, tuned for prompt adherence and visual quality and served for low latency. fal lists the same regular per-second prices for both.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
