Seedance 2.5
#11 in Text-to-VideoByteDance · v2.5 · 8× · tolest 04. Aug. 2026
47
Momentum
Seedance 2.5 is a text-to-video and image-to-video model developed by ByteDance (Volcano Engine/Doubao Team), unveiled on June 23, 2026 at the Volcano Engine FORCE Conference in Beijing. The model is designed to generate continuous 30-second video clips in a single generation pass without stitching together multiple shorter segments, and can process up to 50 multimodal reference inputs (image, video, audio, text). At the time of announcement, Seedance 2.5 was in a global Enterprise Beta phase; a public launch was scheduled for early July 2026, though an exact date and pricing had not yet been officially released. Access is planned through Dreamina, CapCut, Doubao, and the Volcano Engine Ark API.
Momentum-Verloop
07.05.05.08.
Features
| Fine-tuning | Kein Fine-tuning; Steuerung über bis zu 50 multimodale Referenzeingaben (Bild, Video, Audio, Text) sowie regionsbasierte Editierfunktion (Subjekt/Element ersetzen) |
| Max. Videolänge | 30 Sekunden durchgehend in einem Generierungsdurchlauf (Standardmodus); bis zu 180 Sekunden im Beta-Long-Video-Modus laut Dreamina-Produktseite |
| Plattform | Dreamina, CapCut, Doubao, Volcano Engine Model Ark (API) |
| Preis | Offiziell noch nicht veröffentlicht (Stand Ankündigung) |
| Release-Datum | Angekündigt 23. Juni 2026 auf der Volcano Engine FORCE Conference; öffentlicher Launch für Anfang Juli 2026 geplant |
