ByteDance Launches Seedance 2.5: 30s Generation, API Price Hiked Over 50%
智东西 · wechat · 2026-07-31
ByteDance officially released Seedance 2.5, its new generation video generation model. Continuing its unified multimodal audio-visual architecture, the model doubles single-generation duration from 15 to 30 seconds, supporting multi-round extensions up to several minutes to complete coherent narratives.
The model significantly enhances multimodal reference capabilities, accepting up to 30 images, 10 videos, and 10 audio clips in a single prompt, drastically improving consistency in multi-character scenes. Additionally, Seedance 2.5 supports timestamp-based precise editing, green screen rendering, and camera movement adjustments for post-generation modifications.
For enterprise applications, companies like XCMG and XPeng will integrate the model for synthetic data training and industrial simulation. Volcano Engine also announced API pricing at 42 RMB/million tokens (with video input), representing a roughly 50% price increase over the previous generation.
Related event: ByteDance Releases Seedance 2.5: Native 30s Video and Multimodal Control(60 posts)→
More from Multimodal
- MiniMax H3 motion graphics demo impresses, $50K challenge with Picsart opens — egeberkina · 2026-09-16
- After Suno dropped his chords, this user built a MIDI-to-.abc converter so YuE2 preserves them — Saren-WTAKO · 2026-09-16
- Hugging Face team ships 'Building VLMs' book: hands-on guide to vision-language models — andimarafioti · 2026-09-16
- Hugging Face publishes illustrated guide to the 3D generation ecosystem — unofficialmerve · 2026-09-16
- Creator shares 4 imaginary tokens for Midjourney v8.2 to generate metaphorical imagery — LudovicCreator · 2026-09-16
- Zero-cost video generation with a copy-paste Grok prompt that just works — songguoxiansen · 2026-09-16