ByteDance Seedance 2.5 announced at Volcano Engine FORCE June 23 — single native 30-second clip, up to 50 multimodal reference inputs, local re-draw editing changes single frame element without altering rest
ByteDance's Seedance 2.5 was announced June 23 2026 at the Volcano Engine FORCE conference — single native 30-second clip (vs prior 10-15 second baselines), up to 50 multimodal reference inputs in single generation, local re-draw editing that changes one element of a frame without altering the rest. Currently in enterprise beta with early-July public release.
The substantive piece is the three-capability-dimension simultaneous leadership claim: clip duration (30s native), reference control (50 multimodal inputs), local editing (single-element frame modification). Pre-Seedance-2.5 the video-AI capability landscape stratified across vendors with capability specializations. Seedance 2.5's simultaneous leadership across three capability dimensions challenges the stable-stratification pattern.
The competitive read against Veo 3.1's cinematic + integrated-audio leadership + Kling 3.0's cinematic + multi-shot leadership is that Seedance 2.5 may reshuffle H2 2026 video-AI vendor stratification. Three-dimension simultaneous capability leadership (duration + reference control + local editing) is structurally different from single-dimension specialization. Whether the early-July release validates the claims will determine H2 2026 video-AI vendor leadership rotation.
Digital Applied — Seedance 2.5: ByteDance's 30-Second AI Video Model → · Lushbinary — Seedance 2.5 vs Veo, Sora & Kling: AI Video Compared →