Max quality
4K 10-bit
480P / 720P / 1080P / 4K output tiers
Seedance 2.0 is a new generation of AI video generation model launched by ByteDance, focusing on high-quality and long-duration video creation. The current version is the full‑capacity Seedance 2.0 model, which supports real‑life subjects and human faces.
Core highlights:
• Smooth movements: Complex character movements and camera movements are natural and smooth • Precise control: Supports text, image, and pose guidance • Fast generation: Optimized inference, significantly improving generation efficiency
Compared with the previous generation, Seedance 2.0 has comprehensively evolved in terms of picture quality, movement coherence, and semantic understanding, making it suitable for creators to quickly produce professional-grade video content.
$Official Original Pricing Based on Token Billing:
Unconditional discount price: 10% off. Contact us for larger discounts!
The Seedance 2.0 Text to Video model can generate high-quality cinematic videos based on text prompts, featuring native audio synchronization, realistic physical effects, and multi-shot scene transition capabilities.
Seedance 2.0 Reference Material Mode supports 9 reference images, 3 reference videos and 3 audio materials. Material and Scene Limitations Text Prompts: Chinese prompts are recommended to be no more than 500 characters, and English prompts no more than 1000 words. Image Input: Supports jpeg, png, webp, bmp, tiff, gif, heic, heif; each image shall be less than 30 MB, with an aspect ratio ranging from 0.4 to 2.5, and side length between 300px and 6000px. No more than 9 pieces for multimodal reference scenarios. Video Input: Supports mp4, mov; each video lasts 2‑15 seconds, with a maximum of 3 reference videos, and the total duration of all videos shall not exceed 15 seconds; each video shall be less than 50 MB; frame rate ranges from 24 to 60. Audio Input: Supports wav, mp3; each audio clip lasts 2‑15 seconds, with a maximum of 3 reference audio clips, and the total duration of all audio clips shall not exceed 15 seconds; each audio clip shall be less than 15 MB. Audio cannot be uploaded independently; at least one reference image or video shall be included simultaneously. 🔸Note🔸: With video input has a lower unit price due to a different calculation method: No video = Price × Output; With video = Price × (Input + Output).
Seedance 2.0 first and last frame mode can animate any still image into a fluid video clip with consistent character preservation, natural motion, and synchronized audio output.
Seedance 2.0 for Product Teams
From concept to publish-ready clips in minutes.
Create consistent short-form videos for marketing, education, and creator tools with one integration on APIPod.
Generated with Seedance 2.0 · One API away on APIPod
Max quality
4K 10-bit
480P / 720P / 1080P / 4K output tiers
Clip length
4–15s
Coherent multi-shot storytelling
Reference inputs
9 imgs + 3 videos + 3 audios
All-modal references lock identity and style
Native audio
Stereo
Music, SFX, and voice tracks in sync
Built for teams that care about output quality, control, and delivery speed.
Control framing, movement rhythm, and visual constraints with stronger prompt adherence.
Reuse reference packs across modalities to keep identity and style consistent.
Align beat, motion, and pacing for dialogue, brand stories, and rhythm-driven edits.
Iterate weak segments instead of regenerating full videos from scratch.
Maintain more stable motion and interactions in complex scenes.
Use APIPod for unified auth, routing, observability, and cost control in production.
Sourced from the official Seedance launch blog. Press play — original prompts included.
T2V · Complex motion
Synchronized jumps, mid-air spins, and clean landings in pair skating — motion timing and physics stay faithful.
Official prompt
A competitive pair figure-skating performance. Low-angle tracking follows the blades with crisp ice spray. The male skater wobbles mid-spin; his partner steadies him and they flow into a lift. The climax: a synchronized jump combo, straight air posture, decisive landings, perfect audio-visual sync.
R2V · All-modal reference
Combine up to 9 images, 3 videos, and 3 audio clips as references to replicate composition, motion, and style.
Official prompt
A girl breaks the dimensional wall and travels through famous paintings. Realistic texture, 3D high-saturation painterly worlds. She gasps under a swirling starry sky, selfies with the pearl-earring girl, passes between two samurais, paints alongside Van Gogh, then turns to the camera with a sweet smile at sunset. High contrast, cinematic, seamless transitions.
T2V · Multi-shot story
A director's mindset: the model plans its own shot language and completes a full narrative arc within 15 seconds.
Official prompt
A Lunar New Year family album comes alive: portraits animate as the camera pans past — grandparents, parents, and kids each perform a signature gesture. Lanterns and couplets light up as the frame converges into a lively family photo, laughter in sync.
T2V · Stereo sound
Native stereo sound: rain, thunder, and the shockwave of clashing blades land exactly on the frame.
Official prompt
A wuxia duel in a bamboo forest: a white-robed swordsman faces a coir-clad blade fighter. Slow push-in, focus shifting between raindrops and sword hilts — only rain is audible. Thunder cracks, both charge; the clash drops into extreme slow motion, rain rings blasted off the blades, bamboo leaves sliced mid-air.
R2V · Video extension
Extend existing footage with new plot beats while shot language and style stay consistent.
Official prompt
Extend the video: track the orange-robed rider to a tree of orange blossoms; he picks two flowers, dismounts as the camera orbits, and offers them to a white-robed woman on a white horse. Chinese court-lady painting style, 3D, cheerful folk music, shadow-puppet aesthetic in black, white, and orange.
I2V · Commercial
Image-to-video plus a story brief yields a ready-to-ship commercial with captions and voiceover.
Official prompt
The man in the painting nervously peeks out of the frame, grabs a cola, and drinks with pure satisfaction. Footsteps approach — he slips it back as a cowboy walks off with it. End on a top-lit cola close-up against pure black, artistic captions and VO: 'Cola you can't resist!'
Video source: official launch blog at seed.bytedance.com
A practical loop that reduces rework and speeds up content delivery.
Capture objective, scene, camera plan, and style constraints in a structured prompt brief.
Provide keyframes, motion references, and optional audio mood tracks to lock direction early.
Generate short drafts first, evaluate consistency, then promote high-scoring outputs.
Use continuation or targeted edits to improve weak segments while keeping strong shots.
Starting templates you can adapt for real product scenarios.
Brand film teaser
A cinematic close-up of premium coffee beans falling in slow motion, warm tungsten highlights, shallow depth of field, subtle steam, percussion rhythm synchronized with the pour-over shot.
Explainer sequence
Three-shot flow: whiteboard close-up, medium presenter frame, cutaway to animated chart; clean color palette, calm pacing, narration-friendly rhythm.
Game trailer clip
Cyberpunk city rain sequence, aggressive handheld camera, reflective asphalt, bass impacts on transitions, maintain protagonist armor identity across cuts.
Example request
Use APIPod as your unified gateway for routing, billing, retries, and observability.
Parameter names may vary by provider channel and model revision. Use your live APIPod model docs as the source of truth.
curl -X POST https://api.apipod.ai/v1/videos/generations \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d {
"model": "seedance-2.0",
"prompt": "Cinematic night market sequence with wet reflections and synchronized ambient audio.",
"duration": 10,
"resolution": "1080p",
"aspect_ratio": "16:9",
"generate_audio": true
}No. It supports text, image, video, and audio guided generation in one workflow.
Most teams start with short multi-shot clips. Exact limits depend on your enabled channel and policy settings.
Yes. Directed prompts plus reusable references are designed to improve character and style continuity.
APIPod gives you one API layer for authentication, routing, observability, and cost governance across model channels.
Move from experimentation to stable delivery with one integration layer and production-grade monitoring.