Seedance 2.5
Seedance 2.5 is ByteDance's newest video model, launched on 31 July 2026 after a June announcement at Volcano Engine's FORCE conference. Its headline is length: up to 30 seconds in a single pass, reasoning about the whole shot so the subject and lighting hold from first frame to last, with up to 50 multimodal references and local editing. On fal it is billed by the token at about half again the rate of 2.0.
Anvisha Pai, Co-founder & CEO, Voyager
ByteDance · released
›What we asked for
A self-portrait of Seedance 2.5, ByteDance's 30-second video model: a single continuous shot following a red seed as it falls from a tree, lands, sprouts and grows into a sapling in time-lapse, golden hour, natural sound.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 pixels
At a glance
- Price
- About $0.46 a second at 720p on fal, billed by the token: our five-second clips cost $2.31.
- Length
- 4 to 30 seconds in one shot, double 2.0's ceiling.
- Audio
- Native and synchronised, on by default.
- Sizes
- 480p, 720p or 1080p; no 4K on this model.
- References
- Up to 50 images, clips and audio tracks on the reference endpoint.
Other Seedance models: Seedance 2.0. Compare all 2.
Anvisha's take
Seedance 2.5 costs half again as much as 2.0 and, on five-second clips, does not look it. The clips are handsome, the speaking woman is the most naturally lit on the roster, and the title card is crisp. But the turntable clip opened on a black blur before pulling out, the glass in the physics clip lost its water to a haze, the image-to-video job was refused as a copyright risk against our own product photo, and one clip took 26 minutes to come back. For five seconds, I would still start with 2.0. The September 18 supplemental tests below now cover 20 and 30 seconds: both returned at 1280x720, with the red mug and blue apron recognizable at the sampled checkpoints. That adds evidence for long-shot use without erasing the short-shot failures.
Seedance 2.5 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 Seedance 2.5 returned, one try each, nothing re-rolled or retouched. Each tile shows one frame of the clip until you press play. Generated on . The 8 clips took 170 to 1578 seconds each and cost $18.49 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
returned 1280 by 720 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.)
bytedance/seedance-2.5/image-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 5, generate_audio: true
How the cost under each picture was worked out: fal's Seedance 2.5 rate: $0.0214 per 1,000 tokens, tokens = height x width x seconds x 24 / 1024, at the measured 1280x720 for 5 seconds; fal does not return the charge.
Additional workflow tests
Separate from the common benchmark above. These tests examine longer shots, successive edits or specialized controls. One attempt per test, including failures. Costs are estimates from published rates.
20-second continuous workshop shot
Seedance 2.5
One unbroken shot, stable apron, mug and window; the action must reach the final pose.
›What we asked for
One continuous 20-second take in a sunlit pottery workshop, no cuts. A woman wearing a plain blue apron stands at a wooden table with one red mug and a folded white cloth. First third: she picks up the red mug and turns it once in her hands. Middle third: she places it back on the table and wipes beside it with the white cloth. Final third: she folds the cloth, places it to the left of the mug, and looks toward the same window. The camera slowly tracks sideways throughout. Keep the same woman, apron, red mug, table and window. Natural workshop ambience, no music, no text.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 20, generate_audio: true
returned 1280 by 720 pixels
Generated . Cost basis: fal's Seedance 2.5 rate: $0.0214 per 1,000 tokens, tokens = height x width x seconds x 24 / 1024, at the measured 1280x720 for 20 seconds; fal does not return the charge; rate verified 2026-09-18 at https://fal.ai/models/bytedance/seedance-2.5/text-to-video. Original output.
Returned 20.09 seconds at 1280x720. The sampled frames retain the blue apron and red mug and reach a final look toward the window. The framing moves from hands and table to the subject’s face; sampled frames alone do not establish cut-free continuity.
30-second continuous workshop shot
Seedance 2.5
One unbroken shot, stable apron, mug and window; the action must reach the final pose.
›What we asked for
One continuous 30-second take in a sunlit pottery workshop, no cuts. A woman wearing a plain blue apron stands at a wooden table with one red mug and a folded white cloth. First third: she picks up the red mug and turns it once in her hands. Middle third: she places it back on the table and wipes beside it with the white cloth. Final third: she folds the cloth, places it to the left of the mug, and looks toward the same window. The camera slowly tracks sideways throughout. Keep the same woman, apron, red mug, table and window. Natural workshop ambience, no music, no text.
bytedance/seedance-2.5/text-to-video
aspect_ratio: 16:9, resolution: 720p, duration: 30, generate_audio: true
returned 1280 by 720 pixels
Generated . Cost basis: fal's Seedance 2.5 rate: $0.0214 per 1,000 tokens, tokens = height x width x seconds x 24 / 1024, at the measured 1280x720 for 30 seconds; fal does not return the charge; rate verified 2026-09-18 at https://fal.ai/models/bytedance/seedance-2.5/text-to-video. Original output.
Returned 30.09 seconds at 1280x720. The sampled checkpoints keep the apron, mug and workshop recognizable and reach the final window-facing pose. Estimated cost $13.8672; the request returned in about 391 seconds.
Compare Seedance 2.5
Seedance 2.5 pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| fal | Second at 720p, 16:9 $0.0214 per 1,000 tokens at 480p and 720p, tokens = height x width x seconds x 24 / 1024, so $0.4622 a second at 1280 by 720, which is what every clip below cost; fal's page rounds the per-second figure to about $0.47. About $0.22 a second at 480p. | $0.4622 | fal model page, read |
| fal | Five-second 720p clip | $2.31 | fal model page, read |
Is Seedance 2.5 free?
Seedance is free with daily credits inside ByteDance's own apps, Dreamina internationally and Jimeng and Doubao in China. The API, through BytePlus, Volcano Engine and hosts like fal, is paid by the token.
- •Dreamina: free daily credits for images and video; paid plans for more.
- •Jimeng and Doubao: ByteDance's China apps, free with limits.
- •API and hosts: pay per token, no free tier.
Dreamina, read
Where to use Seedance 2.5
- Dreamina · ByteDance's consumer app, free daily credits.
- BytePlus ModelArk · ByteDance's API for international customers.
- fal · Token-billed per clip; text, image and reference endpoints.
- Voyager · Runs inside Voyager, the agent for creative work. Private preview by waitlist.
Voyager
Run Seedance 2.5 inside Voyager
Voyager is an agent for creative work that runs every model in this reference, Seedance 2.5 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 Seedance 2.5
- •Opened on an extreme close-up of the dripper's black rim, so the first second is an abstract before the camera pulls out to the product. See the product turntable clip.
- •Refused: fal returned a 422 saying the generated video was rejected as a potential copyright violation. The starting image was our own Nano Banana Pro product shot of a plain dripper. See the image to video clip.
- •The glass tips over and the ice slides out, but the water turns to a translucent haze rather than a spill. See the physics clip.
- •Waits ranged from 170 seconds to 26 minutes for the same five-second request; budget for the long tail.
Seedance 2.5 limits and API parameters
| Limit | Value |
|---|---|
| Length | 4 to 30 seconds in one pass. fal API reference, read |
| Resolution | 480p, 720p or 1080p; no 4K. fal API reference, read |
›Every setting, for developers
fal (bytedance/seedance-2.5/text-to-video)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| resolution | enum | 480p, 720p, 1080p | 720p |
| duration | enum | auto, 4 to 30 auto lets the model pick a length for the prompt. | auto |
| aspect_ratio | enum | auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16 | auto |
| generate_audio | boolean | true, false | true |
| bitrate_mode | enum | standard, high | standard |
| image_url | URL | Starting frame Image-to-video endpoint; end_image_url sets a closing frame. |
What Seedance 2.5 will not make
ByteDance's models run behind fal's safety checker and fal's partner terms for commercial use, and Seedance 2.0 is the model whose launch drew cease-and-desist letters from Disney and Paramount over recognisable characters and celebrity likenesses, after which ByteDance said it would strengthen its IP safeguards. Expect prompts naming real people, franchises or studios to be refused or degraded. Outputs from Dreamina carry an AI-generated label.
Seedance 2.0 on Wikipedia, including the February 2026 letters, read
Alternatives to Seedance 2.5
Frequently asked questions
What is Seedance 2.5?
Seedance 2.5 is ByteDance's video model launched on 31 July 2026, generating single-shot clips up to 30 seconds with native audio and up to 50 references, at 480p to 1080p.
How much does Seedance 2.5 cost?
On fal, $0.0214 per 1,000 tokens, which works out to $0.46 a second at 1280 by 720, so a five-second clip is $2.31, read on 9 September 2026.
Seedance 2.5 or 2.0?
2.0 unless you need more than 15 seconds in one shot. 2.5 costs half again as much per second and drops 4K. Both pages run the same nine clips.
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.
