Veo 3.1 is coming soon, and Google’s clearly aiming it right at Sora 2 with longer video support (www.techradar.com)

Google’s Veo 3.1 is starting to appear through third‑party AI video services (Higgsfield, Imagine Art, Envato) and brings several production‑grade upgrades that squarely target OpenAI’s Sora 2. Building on Veo’s hallmark lip‑sync and audio alignment, 3.1 reportedly adds native 1080p output, multi‑prompt/multi‑shot support, cinematic presets (drone, pans, tracking, lighting/tone) and much stronger character and scene consistency across shots. Vendors advertise “30 seconds+” clips and leaks and a filmmaker’s post claim the model may handle up to one‑minute videos and a wider range of input image types (photos, illustrations, cartoons). Technically, those changes matter: multi‑prompting lets creators stitch several scene prompts from one source image, camera‑move presets reduce prompt engineering, and tighter frame‑to‑frame consistency (no changing eye color or finger counts) makes longer narrative sequences feasible. Availability via third parties — not yet in Gemini — means rapid experimentation but also raises the same creative, legal and ethical questions already dogging Sora (rights, likenesses, and attribution). If Veo 3.1 delivers on quality and length, it will heighten competition in AI video, accelerating tools that push short‑form generative video toward more cinematic, multi‑shot storytelling.