How do I get sound in my videos?
Most current video models generate native audio. If a clip came out silent, Video Audio can add synchronized sound afterwards.
Generate it with the video
Most current models produce native audio as part of the generation: Gemini Omni Flash, the Kling O3 family, Kling 3 and 3 Pro, Kling 2.6, the Seedance 2.x family, Seedance 1.5 Pro, Hailuo 03, WAN 2.6, LTX 2, Sora 2 and VEO 3.1. Look for the audio toggle in the settings before you generate.
On Gemini Omni Flash you control the audio directly in your prompt — describe the sound effects or the dialogue you want and it will produce them.
Audio can change the price
On some models audio is a separate rate. Kling 3 is 20 credits per second without audio and 30 with it; Kling 3 Pro is 25 against 40. VEO 3.1 is 225 per clip without audio and 338 with. The cost shown before you generate always reflects your current toggle.
Add sound to a clip that came out silent
Video Audio takes an existing silent video and generates synchronized sound for it using MMAudio v2. Describe the audio you want, or use the "AI Understand Video" option to have it suggested from the footage. Uploads up to 100 MB.
Add music or a voiceover instead
Generate a track in the Music Generator or a voiceover in Voice Generation, then lay both over your video in the Video Editor. Timeline editing and export are free, so this costs only the generation itself.
Last reviewed against the live product in August 2026.