Computational Model Studio · Model release
Seedance 2.5
Generate and extend up to 30-second audio-video clips from multimodal creative direction.
Developer: ByteDance Seed · Last verified:
What it is designed to do
Generate and extend up to 30-second audio-video clips from multimodal creative direction.
- Primary contribution
- generation transformation
- Input
- text, image, audio, video
- Output
- video, audio (generated audio-video clip)
Documented uses: Generate and extend up to 30-second audio-video clips from multimodal creative direction.
Access and deployment
- Seedance 2.5 web experience — available. Select Seedance 2.5 in Jimeng Web or Doubao Pro where available. Offering documentation Complete publisher web service terms were not reviewed for this offering.
- provider hosted: Select Seedance 2.5 in the publisher’s documented web experience. Limits: A public API model ID or local weight artifact is not inferred from this web selector..
Technical profile and review notes
- Construction
- learned
- Representation
- neural
- Normalized architecture tags
- No normalized architecture tag assigned
- Lifecycle
- active
Inclusion rationale: ByteDance documents a current long-form, reference-based audio-video model with web access while API access is pending.
Limitations
- The publisher’s described output is recorded as the exact media form shown; surrounding editing tools or orchestration are not attributed to this model.
- The documented web model selector does not establish an exact public API model ID or downloadable weights.
- The publisher states API access is coming soon; only the documented web access is recorded.
Sources and related Studio records
Sources
Connections
- In areaComputational Models
- Supports functionGenerate video
- OfferingSeedance 2.5 web experience