AI Video Generator - AI Studio

This update is about two things: putting the world's best AI models in your pocket, and giving you new ways to make something with them. Meet the new lineup. Gemini Omni 1.1 Flash takes our top spot: video with natively synced sound, now up to 10 seconds. Wan 3.0 arrives speaking full audio and stretching to 30-second videos. MiniMax H3 Max joins with impossibly smooth motion. And three new image models move into the studio; Muse by Meta (completely free for subscribers), Microsoft's MAI Image 2.5, and Imagine 2.0 from xAI. Can't decide who should make your next video? Neither could we, so we built Model Ranking, right inside the model picker: real quality scores from arena.ai's community battles, so you always know who's on top. And there's a new tool in the studio: Text Brush. Drag your finger across a photo and your words follow the line, bending with every curve you draw. Every font on your phone is in there, with weight, color, size and opacity to match, then save straight to your library. Creating got simpler too. Reference to Video is now your starting point, every mode is named for what it actually does, and the Model Fallback switch in Settings, now covering every single video model we offer, puts you in charge of whether we auto-rescue a failed generation with a backup. Director picked up some polish as well: the photos you choose now travel with the message you sent them in. Go make something amazing. Thank you for creating with AI Studio. Every video you make, every review you leave, and every idea you send us shapes what we build next; this update is full of things you asked for. We read everything. Keep it coming.