VideoByteDance · Volcengine Ark
Seedance 2.0
Cinematic, multi-modal AI video with multi-shot consistency.
Seedance 2.0 generates cinematic video from text, images, and references — with coherent multi-shot storytelling, stable characters across scenes, and synced audio. Built for creators who want directorial control, not random clips.
Generated with Seedance 2.0
Capabilities
- Text → Video
- Image → Video
- First & Last Frame → Video
- Multi-Reference / Omni Reference
- Synced audio & lip-match
- Multi-shot character consistency
At a glance
- Output
- Up to 2K
- Inputs
- Text · Image · Video · Audio
- Clip length
- Up to 15s
- Best for
- Hero & narrative video
Sample outputs
Official sample outputs from Seedance 2.0.
Use via API
For developersOne endpoint, one key — just switch the model parameter to call any model. No need to integrate ByteDance directly, manage separate billing, or maintain four SDKs.
curl -X POST https://pixmo.ai/api/v1/jobs/createTask \
-H "Authorization: Bearer $PIXMO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "sd2",
"inputs": { "prompt": "your prompt here" }
}'
# returns { "taskId": "..." } — then poll:
curl "https://pixmo.ai/api/v1/jobs/queryTask?taskId=TASK_ID" \
-H "Authorization: Bearer $PIXMO_API_KEY"
