Seedance 2.0 nails "possessed fighting an invisible enemy" with no edits
gen_ericai · x · 2026-09-28
genericai shares a Seedance 2.0 generated video of a possessed character fighting an invisible enemy, claiming no edits and no image reference—pure prompt. A showcase of the video model's ability to convey unseen opponents and dramatic acting.
More from Multimodal
- Revisiting a DALL-E beatbox game proto: what scaled with models vs. got replaced — Flomerboy · 2026-09-28
- Qwen Image 2.1 LoRA boom: most finetunes add little over native ability — linoy_tsaban · 2026-09-28
- Indie musician uses AI to finally bring a 3-year-old unreleased song to life — itsmnjn · 2026-09-28
- Kokoro-82M: tiny open-source TTS model runs on CPU with quality of far larger models — solyarisoftware · 2026-09-28
- MiniMax H3 ecosystem roundup: seam fix merged, new LoRAs and ComfyUI workflow tools — optimisticalish · 2026-09-28
- Vampire fantasy movie trailer Erebus 2 made with Seedance AI — Interesting_Tone6532 · 2026-09-28