Include five elements: scene description (who, what, where), camera direction (angle, movement), lighting (sources, color temperature), mood (emotional tone), and audio direction (sounds, dialogue, music). Seedance 2 understands all five and uses them to shape both the video and the four audio layers. Be specific rather than generic - 'neon reflections on wet pavement' works better than 'cool lighting'.

Yes. Your text prompt directs all four audio layers: dialogue (with phoneme-level lip sync), foley effects, ambient atmosphere, and background music. If you describe rain, you get rain sounds. If you write character dialogue, you get lip-synced speech. If you mention music, you get a matching score. The audio is created in the same pass as the video - no post-production needed.

Six options: 16:9 (YouTube/cinema), 9:16 (TikTok/Reels), 1:1 (Instagram), 4:3 (presentations), 3:4 (Pinterest/portrait), and 21:9 (ultra-wide cinematic). All output at 30fps. You can specify the aspect ratio in your prompt or let TwoShot's assistant choose based on your intended platform.

Seedance 2 creates clips from 4 to 15 seconds. Two modes are available: Fast mode for quick iteration (lower cost, faster delivery) and Standard mode for maximum quality. For longer content, you can create multiple clips and combine them in TwoShot's studio timeline - the character consistency feature ensures your characters look the same across all clips.

Yes. Write the dialogue directly in your prompt and Seedance 2 will create lip-synced speech. The model supports 8+ languages including English, Mandarin, Korean, Japanese, Spanish, and Indonesian. Phoneme-level lip sync means mouth movements match the actual sounds, not just generic mouth opening and closing.

Fast mode delivers results in seconds. Standard mode takes under 60 seconds. Both modes are available with free credits on signup - no queue, no waitlist. You can iterate on prompts quickly using Fast mode, then switch to Standard for the final version.

Absolutely. You can upload reference images alongside your text prompt. For example: upload a character portrait, then write a text prompt describing a scene for that character. Seedance 2 uses the image for visual identity and the text for scene direction, camera movement, and audio. This hybrid approach gives you the best of both input modes.