Seedance 2.5 Flash
ByteDance's fast, low-cost video model — 30-second clips at 480p/720p with native audio and up to 20 reference assets per generation.
# image_url: the first frame, as a URL or a data: URL
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-2-5-flash-image-to-video-basic",
"prompt": "The camera slowly pushes in as mist drifts across the scene",
"image_url": "https://example.com/start-frame.jpg",
"duration": "5s",
"resolution": "720p"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Overview
What is Seedance 2.5 Flash
Seedance 2.5 Flash is ByteDance's accelerated, cost-reduced video generation model from the Doubao team, released November 12, 2026. It renders clips up to 30 seconds at 480p or 720p with native audio, accepting a text prompt, a still image, or up to 20 reference assets — at roughly twice the speed of flagship Seedance 2.5.
Using it anonymously on Venice
On Venice, Seedance 2.5 Flash runs under the anonymized privacy tier — your prompts and reference images are not stored, profiled, or used for training, and nothing ties a clip back to a personal history. You pay per clip in credits instead of a platform subscription, with all three modes (text-to-video, image-to-video, reference-to-video) available. Venice listed the model on November 14, 2026, the same week BytePlus announced it.
Specifications
Datasheet
- Maker
- ByteDance (Doubao team), distributed via BytePlus / Dreamina
- Modality
- Text-to-video, image-to-video, reference-to-video
- Open weights
- No
- License
- Proprietary
- Clip lengths
- 4s – 30s
- Resolutions
- 720p, 480p
- Mode
- image-to-video
- Audio
- Yes
- Prompt limit
- 15,000 chars
- Input images
- Up to 10 reference images (accepted formats and size limits not published)
- Released
- November 12, 2026
- Architecture
- Diffusion Transformer (DiT), per Volcengine launch disclosures
- Privacy on Venice
- Anonymized — prompts not stored
- Available on Venice since
- Nov 2026
Assessment
Strengths and limitations
- Longest single-generation clips in its class: up to 30 seconds in one pass, double the 15–16 second cap of most rival video models on Venice.
- Native audio generated in the same pass as the video, so dialogue, ambience, and sound effects arrive with the clip rather than in a separate step.
- Rich multi-reference input: up to 10 images, 5 videos, and 5 audio clips per generation keep characters, products, and style consistent across shots.
- First-and-last-frame control lets you define exactly where a shot starts and ends.
- Positioned as the fast, cheap tier: BytePlus claims roughly twice the generation speed of Seedance 2.5 at about 34% lower cost per second, with output described as the closest to the flagship in the Seedance family.
- Capped at 480p and 720p: 1080p and 4K output are explicitly not supported; those resolutions stay with the full Seedance 2.5.
- Closed and proprietary: no open weights, so it cannot be self-hosted or fine-tuned.
- The 2× speed claim is BytePlus's own marketing figure; no independent timing benchmarks were available at launch.
- Third-party testing of the Seedance family reports that it declines real-person photos as opening frames, unlike some rivals — worth testing for your specific use case.
- Because it is a Flash tier, fine detail and complex motion may trail the full Seedance 2.5 for final-render work.
Use cases
What it is good for
- 01Long-form social content — weekly feeds, stories, and product moments where a full 30-second take renders in one pass.
- 02Ad hook testing: render several openings side by side on a fast, cheap tier before committing to a final render.
- 03Animating still images into motion for product demos and marketing creative.
- 04Multi-shot sequences with a consistent character or product, locked in via reference images.
- 05Drafting and pacing shots on Flash, then re-rendering the keeper on full Seedance 2.5.
Prompting
Getting better results
Use the reference-to-video variant (seedance-2-5-flash-reference-to-video-basic) with up to 10 images to lock a character, product, or style across the whole clip.
Supply a first frame and a last frame when you need the motion arc to land on an exact ending — the model supports first-and-last-frame control.
Draft at 480p to iterate cheaply from $0.34 per clip, then re-render the winning prompt at 720p.
Write dialogue and sound cues directly into the prompt — audio is generated in the same pass, so described speech and ambience get rendered natively.
For 30-second clips, use the generous 15,000-character prompt budget to lay out the sequence beat by beat ("0–10s: …, 10–20s: …") instead of one vague sentence.
Name camera moves explicitly ("slow dolly in", "handheld follow") — motion-physics adherence is a stated strength of the Seedance line, and explicit direction exploits it.
Samples
Sample outputs
Generated on Venice with our standard prompt suite — the same prompts we run through every model of this type, so you can judge it like-for-like.
Slow aerial drone shot gliding over a misty mountain valley at golden hour, sunlight piercing clouds onto a winding river, ancient pine forests on either side, ultra-smooth motion, professional color grading, atmospheric haze, 4K
Calm ocean waves rolling onto a black-sand beach at sunrise, golden light on wet sand, foam dissolving into the shore, a single silhouetted figure at the waterline, smooth continuous forward push, serene cinematic atmosphere
A slow tracking shot through a rain-soaked Tokyo street at night, neon reflecting in puddles, steam rising from a food cart, a person with a translucent umbrella, shallow depth of field, teal-and-orange grade, smooth steady camera
Alternatives
How it compares
| Model | Best for | Max duration | Max resolution | Native audio |
|---|---|---|---|---|
| Seedance 2.5 Flash | Long clips with references & audio | 30s | 720p | Yes |
| Grok Imagine 1.5 | Fast, sharp social clips | 15s | 1080p | Yes |
| Wan 2.7 Enhanced | Open-weights workflows | 15s | 1080p | Yes |
| MiniMax H3 Max | Cinematic 1080p shots | 15s | 1080P | Yes |
Seedance 2.5 Flash is the right pick when the clip has to run long — nothing else on this table renders 30 seconds in one pass, and it does it with native audio and multi-reference consistency. If 1080p sharpness matters more than duration, Grok Imagine 1.5 or MiniMax H3 Max is the better call.
API
Call it from your code
Venice exposes this model through the REST API. Queue a generation with the model id and the inputs it needs.
# image_url: the first frame, as a URL or a data: URL
curl https://api.venice.ai/api/v1/video/queue \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedance-2-5-flash-image-to-video-basic",
"prompt": "The camera slowly pushes in as mist drifts across the scene",
"image_url": "https://example.com/start-frame.jpg",
"duration": "5s",
"resolution": "720p"
}'
# Use the returned queue_id with https://api.venice.ai/api/v1/video/retrieve.
# Call /video/complete after downloading if needed.Pricing
What it costs on Venice
Pay per clip on Venice — price scales with resolution and duration (4s–30s), from $0.34.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
FAQ
Frequently asked questions
Seedance 2.5 Flash is ByteDance's accelerated, cost-reduced video generation model from the Doubao team, announced November 12, 2026. It generates clips up to 30 seconds at 480p or 720p with native audio, and accepts text prompts, still images, or up to 20 reference assets (10 images, 5 videos, 5 audio clips) per generation.
Venice bills per clip, with the price scaling by resolution and duration: a 4-second 480p clip starts at $0.34 and a 4-second 720p clip is $0.76, rising with clip length up to 30 seconds. There is no subscription required — you pay in credits per generation.
No. Seedance 2.5 Flash is closed and proprietary to ByteDance, with no open weights available, so it cannot be self-hosted or fine-tuned. If open weights matter to you, Wan 2.7 and Wan 2.7 Enhanced on Venice are open-weight video alternatives.
Seedance 2.5 Flash wins on duration and references: 30-second clips with up to 20 reference assets and first-and-last-frame control. Grok Imagine 1.5 wins on resolution, outputting 1080p against Seedance 2.5 Flash's 720p cap, but tops out at 15 seconds. Choose Seedance for long, consistent sequences; Grok for shorter, sharper clips.
480p and 720p only. 1080p and 4K are explicitly not supported on the Flash tier — those resolutions remain exclusive to the full Seedance 2.5 model.
Yes. The image-to-video variant (seedance-2-5-flash-image-to-video-basic) animates a still image into motion, and the family also includes text-to-video (seedance-2-5-flash-text-to-video-basic) and reference-to-video (seedance-2-5-flash-reference-to-video-basic) for subject consistency. All three run 4–30 second clips at 480p or 720p with audio.
Between 4 and 30 seconds per generation, selectable in one-second increments. That 30-second ceiling is double the 15–16 second cap of most competing video models and is one of the model's main selling points.
Yes. Audio is generated natively in the same pass as the video — dialogue, ambience, and sound effects come with the clip rather than requiring a separate audio tool. It also accepts up to 5 audio tracks as references to steer the soundtrack.
Venice runs Seedance 2.5 Flash under its anonymized privacy tier: prompts and reference assets are not stored, profiled, or used for training, and generations are not tied to a personal history. You can use it in the Venice app or via the Venice API without building an account-linked generation record.
Use Seedance 2.5 Flash anonymously
Venice does not store your prompts. Chat history stays in your browser.