Doubao SeedDance 2.0 Fast — faster, lower-cost variant of SeedDance 2.0 for iterative video drafting.
Doubao Seedance 2.0 Fast is ByteDance's faster, lower-cost variant of Seedance 2.0, designed for iterative video drafting and rapid prototyping. Built on the same unified multimodal architecture as the standard Seedance 2.0 — supporting text-to-video, image-to-video, multimodal references (up to 9 images, 3 video clips, 3 audio clips), and native audio generation — Fast trades resolution and fine-grained quality for significantly reduced generation cost and slightly shorter latency.
Like its flagship sibling, Seedance 2.0 Fast generates up to 15-second videos at 24 fps with native audio (dialogue, sound effects, ambient sound, and music), supports first-and-last-frame control for smooth looping, and handles camera motion through natural language prompts. The key trade-off is output resolution: the Fast variant caps at 720p versus the standard model's 1080p, and its prompt adherence and temporal consistency scores are slightly lower in independent benchmarks. For iterating on storyboards, testing scene compositions, and budget-conscious rapid prototyping, Fast is the right choice — final renders should use the standard Seedance 2.0 tier.
| Type | Rate |
|---|---|
| Video | $0.1296 / second |
request is price per second for 720x1280 or 1280x720 outputs — the vendor's own per-second rate for this catalog configuration (fast tier)
| Release date | 2026-02-12seed.bytedance.com ↗ | verified |
| API release date | 2026-04-02apidog.com ↗ | verified |
| Max output resolution | 720pmaxvideoai.com ↗ | verified |
| Max duration | 15 secondsapidog.com ↗ | verified |
| Output format | MP4 @ 24 fpswww.volcengine.com ↗ | verified |
| Aspect ratios | Auto, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16maxvideoai.com ↗ | verified |
| Native audio | Dialogue, sound effects, ambient sound, music, lip sync (8+ languages)seed.bytedance.com ↗ | verified |
| Model version | doubao-seedance-2-0-fast-260128www.volcengine.com ↗ | verified |
| Architecture | Unified multimodal audio-video joint generation (text + image + video + audio inputs)seed.bytedance.com ↗ | verified |
| License | Commercial use permitted via Volcengine Ark / TheRouter APIwww.volcengine.com ↗ | to verify |
| Benchmark | Distribution | Score | Source |
|---|---|---|---|
MaxVideoAI Quality Score Aggregate quality score across 11 criteria (prompt adherence, visual quality, motion realism, etc.); scored against Seedance 2.0 (8.5/10). | 7.8/10 | maxvideoai.com ↗ | |
Motion Realism Motion smoothness and physics plausibility (Seedance 2.0: 8.7/10). Fast variant retains strong motion quality despite lower resolution. | 8.1/10 | maxvideoai.com ↗ | |
Prompt Adherence Instruction-following quality (Seedance 2.0: 8.6/10). The gap is small enough for pre-visualization but noticeable in precise constraint scenarios. | 8.0/10 | maxvideoai.com ↗ |
Use the global api.therouter.ai endpoint shown below for new integrations; the legacy China accelerated endpoint is retired.
curl https://api.therouter.ai/v1/videos -H "Content-Type: application/json" -H "Authorization: Bearer $THE_ROUTER_API_KEY" -d '{
"model": "doubao/doubao-seedance-2-0-fast",
"prompt": "A slow tracking shot through a neon-lit city street",
"duration": 8
}'Generate videos using TheRouter's POST /v1/videos endpoint. Submit an async task, then poll GET /v1/videos/:id for the result.
curl https://api.therouter.ai/v1/videos \
-H "Authorization: Bearer $THEROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "doubao/doubao-seedance-2-0-fast",
"prompt": "A golden retriever running through a sunlit wheat field, wide cinematic tracking shot, warm golden hour lighting",
"seconds": 5,
"size": "720p",
"aspect_ratio": "16:9"
}'ByteDance released the official Seedance 2.0 API through Volcengine Ark, enabling programmatic video generation with the async submit-poll-download pattern. Both the standard model and Fast variant are available via API. Through TheRouter, you can access both from a single OpenAI-compatible endpoint (POST /v1/videos) with unified billing.
ByteDance's Seed team announced Seedance 2.0, a next-generation video generation model with a unified audio-video architecture. Key advances include: multimodal input (text, image, video, audio), first-and-last-frame control, camera motion via natural language, and native audio output including lip sync. The Fast variant launched alongside for cost-effective iterative drafting at 720p.
Seedance 2.0 Fast is the lower-cost, slightly faster variant capped at 720p output, trading about 8% quality (based on aggregate benchmarks) for roughly 30% lower per-video cost. It shares the same multimodal architecture, supports the same input types, and produces the same 15s/24fps output format. Use Fast for iterative drafting and standard tier for final production renders.
Yes — TheRouter exposes Seedance 2.0 Fast via POST /v1/videos (accessible through client.videos.create() in the OpenAI SDK) and GET /v1/videos/:id (client.videos.retrieve()). Set the baseURL to https://api.therouter.ai/v1 and pass the model id doubao/doubao-seedance-2-0-fast. Video generation is an async task — it returns immediately with a task ID, and you poll the GET endpoint for the completed video URL.
| source | URL | retrieved | |
|---|---|---|---|
| Release date | seed.bytedance.com ↗ | 2026-05-26 | verified |
| API release date | apidog.com ↗ | 2026-05-26 | verified |
| Max output resolution | maxvideoai.com ↗ | 2026-05-26 | verified |
| Max duration | apidog.com ↗ | 2026-05-26 | verified |
| Output format | www.volcengine.com ↗ | 2026-05-26 | verified |
| Aspect ratios | maxvideoai.com ↗ | 2026-05-26 | verified |
| Native audio | seed.bytedance.com ↗ | 2026-05-26 | verified |
| Model version | www.volcengine.com ↗ | 2026-05-26 | verified |
| Architecture | seed.bytedance.com ↗ | 2026-05-26 | verified |
| License | www.volcengine.com ↗ | 2026-05-26 | to verify |
| MaxVideoAI Quality Score | maxvideoai.com ↗ | 2026-05-26 | to verify |
| Motion Realism | maxvideoai.com ↗ | 2026-05-26 | to verify |
| Prompt Adherence | maxvideoai.com ↗ | 2026-05-26 | to verify |
| Seedance 2.0 API launches on Volcengine Ark — async task pattern for programmatic access | apidog.com ↗ | 2026-05-26 | verified |
| ByteDance launches Seedance 2.0 — unified multimodal video generation with native audio | seed.bytedance.com ↗ | 2026-05-26 | verified |
| What is the main difference between Seedance 2.0 Fast and the standard Seedance 2.0? | maxvideoai.com ↗ | 2026-05-26 | to verify |
| Can I use Seedance 2.0 Fast through the standard OpenAI SDK? | apidog.com ↗ | 2026-05-26 | to verify |