Google Cuts Video AI Prices and Ships Veo 3.1 as the Model Race for Video Heats Up
Google launched Veo 3.1 with richer audio, cinematic style controls, and a new Lite tier — then cut prices on Veo 3 and Veo 3 Fast simultaneously, compressing the cost of AI video generation across the board as competition with Sora's successors intensifies.
Original sourceGoogle launched Veo 3.1 on April 7, 2026, introducing new creative capabilities — richer native audio generation, reference image guidance, video extension, and transition generation — while simultaneously cutting prices across the entire Veo model family.
The new pricing structure: Veo 3 drops to $0.40/second, Veo 3 Fast to $0.15/second, with Veo 3.1 Lite coming in below 50% of Fast tier costs. For 1080p output via the Fast tier, pricing lands around $0.12/second; 4K is $0.30/second. These are significant cuts from the earlier pricing that made Veo 3 a premium-only feature.
Timing is notable. Reports emerged the same week that OpenAI shut down Sora, citing unsustainable compute costs (estimates put it at $15M+/day at scale). Google's decision to cut video AI prices while shipping new capabilities suggests the company is betting that aggressive pricing — not just model quality — will determine who owns the AI video market.
The Veo 3.1 Lite tier is the new story for developers. At sub-$0.07/second for standard resolution, it makes short-form video generation financially viable for consumer apps, social platforms, and indie developers who couldn't justify the previous cost structure. The model supports text-to-video, image-to-video, and video-to-video — all via the Gemini API and Vertex AI.
For the broader AI video market, the pressure is real. Runway ML, Kling, Pika, and Sora's successors are all competing in a space where Google has infrastructure and distribution advantages. This price drop is a signal that Google intends to use those advantages aggressively.
Panel Takes
The Builder
Developer Perspective
“Veo 3.1 Lite at under $0.07/second changes the economics for consumer video apps. I can now build features that generate 10-second clips without the per-user cost being a business model blocker. The reference image guidance in 3.1 is also genuinely useful for maintaining visual consistency across generated clips.”
The Skeptic
Reality Check
“Price cuts in AI video are great for developers but concerning for the market. If Google is the only player who can sustain loss-leader pricing while Sora gets shut down from compute costs, we're heading toward a duopoly that'll reprice aggressively once competition is weeded out. Low prices today don't mean low prices in 2028.”
The Futurist
Big Picture
“The commoditization of video generation is happening faster than most expected. When you can generate a second of 1080p AI video for $0.12, the question stops being 'can we afford to use AI video' and starts being 'what do we do now that video is effectively free to produce.' Creative work is being fundamentally repriced.”