veo 4 is a cinematic AI video generator that transforms text prompts or reference images into polished 4K clips. It focuses on directing-quality output, supporting camera language such as dolly, crane, rack focus, and whip pan. The platform generates synchronized dialogue, Foley, ambience, and lip-sync directly in the video, reducing the need for post-production. It also keeps character identity, wardrobe, and scene continuity consistent across multiple shots. Additional controls let users adjust aspect ratio, motion intensity, duration, and focal points. With prompt galleries, image-to-video support, commercial usage options, and fast rendering, veo 4 is positioned as a practical tool for creators who need film-like AI video production.

Lyria 3 is a consumer-facing music generation tool integrated into the Gemini app that produces short-form (≈30s) songs with vocals, lyrics, and cover art. It accepts multimodal inputs—text plus images or video—to steer mood, genre, tempo, and vocal character. The system auto-generates lyrics from prompt intent, yields coherent song sections (intro, verse, chorus), and includes SynthID watermarking for provenance. While Lyria 2 remains the enterprise API offering via Vertex AI, Lyria 3 focuses on streamlined creator workflows, fast iteration, and mobile/web access with tiered free and paid limits.