Introduction
Seedance 2.0 Text-to-Video turns a written scene into a video with optional generated sound. Use it to explore a setting, performance or camera idea before making source images. Describe what the viewer should see and hear; this mode does not require an opening photograph.
The model is developed by ByteDance and was introduced in its February 12, 2026 launch article.
Choose the starting point for your Seedance 2.0 sequence
| Starting material | Mode |
|---|---|
| A script or scene description | Seedance 2.0 Text to Video |
| An approved opening still | Seedance 2.0 Image to Video |
| Separate appearance, motion or audio references | Seedance 2.0 Reference to Video |
Key features & highlights
Multimodal creative direction
Seedance 2.0 brings text, images, video and audio into one creative process. References can guide a character's appearance, a scene's composition, movement, camera language or sound, rather than serving only as pictures to animate.
Complex motion and interaction
The model is designed to handle sequences involving several subjects, connected actions and interaction with the environment. Its motion modeling covers both larger movements and smaller details such as gestures, facial expressions and moving fabric.
Audio and video generated together
Speech, environmental sound, effects and music can be generated alongside the picture. This joint approach lets sound participate in the scene's rhythm instead of being treated solely as a separate editing step.
Storytelling and camera control
Natural-language direction can describe a sequence of shots, changes in viewpoint and the progression of a performance. The model family also supports video editing and continuation; those operations require a compatible video-input workflow, not text-only or first-frame generation.
Request parameters
| Parameter | Required | Description |
|---|---|---|
prompt | Yes | The only source of what appears on screen. Order it as subject, action, camera, light, sound, one idea per sentence. |
aspect_ratio | No | Output framing, |
resolution | No | Output resolution, 720p by default and reaching 4k; no upscaled tier is offered on this endpoint. It sets a rate tier, and higher tiers mean bigger files and longer renders. |
duration_seconds | No | Length in whole seconds, 4 to 15, defaulting to 5, and the billed quantity. |
generate_audio | No | Whether a native soundtrack is rendered in sync with the picture. On by default; switch it off when you need picture-only material to score or dub later. Default: |
enable_web_search | No | Off by default. Switched on, the model may search the web for information related to your prompt before it renders. Default: |
seed | No | Integer seed; omit it and each run draws its own. Fix it with everything else held constant to judge what a prompt edit actually changed. |
return_last_frame | No | Also returns the clip’s final frame as a separate image next to the video. Send it to image-to-video as the next opening frame to chain shots without exporting a still yourself. Default: |
bitrate_mode | No | Encoding bitrate of the delivered file: |
watermark | No | Adds the model’s own visible "AI generated" watermark to the video. Off by default; switch it on when the place you publish requires AI-made footage to be labelled. Default: |
Pricing
| Specification | Price |
|---|---|
| $0.07 / second |
| $0.152 / second |
| $0.374 / second |
| $0.778 / second |
| $0.114 / second |
| $0.246 / second |
| $0.612 / second |
| $1.248 / second |
Use cases
Pitch a location before scouting it
Describe an invented bookshop, station or interior and the camera's path through it. Use the clip to discuss atmosphere and framing; review whether the space and movement communicate the intended idea.
Explore a short visual story
Write an event with a clear change, such as an empty market stall becoming ready to open. This is concept footage, not a reproduction of a supplied location. Check that the beginning and ending read clearly without explanatory captions.
Sketch an audio-led moment
Build a scene around a doorbell, footsteps or a brief spoken line. Specify the sound source and the visible reaction. Review whether the action and sound make sense together before using the clip in an edit.
Prompt guide & pro tips
How to write a Seedance 2.0 text-to-video prompt
Build the brief in this order: subject, action, setting, light, camera, visual style and sound. Describe observable actions rather than relying on words such as “cinematic” or “emotional” to supply the story.
- Identify the subject with a few stable details.
- Divide the action into numbered shots, each with one main camera movement.
- Put dialogue or sound beside the action it accompanies.
- Finish with shared style instructions and any text or music exclusions.
Choose duration_seconds of 10 for this example:
A florist in a green apron opens a small street stall at dawn.
Naturalistic photography, soft morning light, muted colors.
Shot 1: A fixed wide shot. The florist lifts the stall shutter, revealing rows of flower buckets.
Shot 2: Cut to a medium side view. The florist sets a bunch of yellow flowers in the front bucket and straightens the stems.
Shot 3: A slow push toward the flowers as the florist steps back and smiles.
Sound: the shutter rattles, paper rustles and distant traffic passes.
No dialogue, background music, captions or added logos.
Use shot order rather than exact timestamps
For Seedance 2.0, use “Shot 1,” “Shot 2” and “Shot 3” to establish order. Exact second-by-second timing is unreliable in this version. If an action disappears, reduce the number of events or give the sequence more time within the supported duration.
Make motion and emotion visible
Replace “the florist is relieved” with a visible performance: the shoulders relax, the florist exhales and smiles. Specify direction and pace for movement, such as “slowly lifts the shutter,” rather than adding more quality adjectives.
Keep camera and sound instructions clear
Avoid asking for an orbit, a zoom and a sideways track in the same shot. Name the speaker when adding dialogue, and keep dialogue in one language unless a language change is intentional. If unwanted captions appear, explicitly request no subtitles; that instruction reduces ambiguity but does not guarantee a text-free result.
Content boundaries
Seedance 2.0 · Text-to-Video — Uncensored. SpicyAPI adds no platform-side content filter. Model restrictions still apply, and results are not guaranteed. You are responsible for complying with our content policy.
API quick start
# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
-H "Authorization: Bearer $SPICY_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
-d '{
"model": "bytedance/seedance-2.0/text-to-video",
"input": {
"prompt": "A slow aerial shot drifting over a misty pine forest at sunrise",
"resolution": "720p",
"aspect_ratio": "adaptive",
"bitrate_mode": "standard",
"generate_audio": true,
"duration_seconds": 5
}
}'
# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
-H "Authorization: Bearer $SPICY_API_KEY"
See the API reference for authentication, the complete schema, polling, and error handling.
