Google launches Veo 3.1 Lite, slashing AI video generation costs as OpenAI exits market

Google Doubles Down on AI Video Generation with Budget-Friendly Model

Google launched Veo 3.1 Lite on Tuesday, positioning it as the company's most cost-effective AI video generator to date 1

1

. The new model costs less than 50% of Veo 3.1 Fast while maintaining the same generation speed, with pricing starting at five cents per second for 720p video with audio 2

2

. This represents a significant shift in Google's strategy to make AI video generation accessible for developers building high-volume video applications without breaking the bank.

The timing carries weight. Just days before Google's announcement, OpenAI shut down its Sora video generation app, reportedly after burning through $15 million per day 4

4

. While OpenAI pivoted away from consumer-facing video AI, Google moved aggressively in the opposite direction, expanding its Veo product family and reaffirming that "video's here to stay" 3

3

.

Pricing Structure Targets Developer Adoption

Veo 3.1 Lite is available exclusively through the paid tier of the Gemini API and Google AI Studio, not through the consumer-facing Gemini app 2

2

. Developers pay five cents per second for 720p video resolution and eight cents per second for 1080p output. To put this in context, the original Veo 3.1 cost around $0.40 per second, while Veo 3.1 Fast ran $0.15 per second 4

4

. The lower pricing makes scalable video output financially viable for smaller creators and startups building video features into their applications.

The model supports both text-to-video and image-to-video generation in landscape (16:9) and portrait (9:16) aspect ratios 3

3

. Developers can customize video duration at 4, 6, or 8 seconds, with API costs adjusting accordingly 1

1

. While the model doesn't support 4K resolution like its premium siblings, this trade-off enables the aggressive pricing that targets high-volume use cases.

Additional Price Cuts Across the Veo Lineup

Google isn't stopping with Veo 3.1 Lite. Starting April 7, Veo 3.1 Fast will see price reductions across all resolution tiers: 720p drops to 10 cents per second (down five cents), 1080p falls to 15 cents per second (down three cents), and 4K video decreases to 35 cents per second (a five-cent reduction) 2

2

. This coordinated pricing strategy signals Google's infrastructure play to establish video generation as a standard component rather than a premium feature.

Competitive Pressure from Chinese AI Companies

The cost reductions address a growing competitive threat. Chinese companies like Kuaishou's Kling AI have offered comparable video generation at lower prices than Google's premium tiers, while Tencent's Hunyuan Video released an open-source model for free during OpenAI's 2024 Sora launch 4

4

. These competitors have been winning the economics argument, forcing Western AI companies to reconsider their pricing models. Google's response targets the middle ground: developers who need to ship video features at scale without hemorrhaging credits on every iteration.

Integration Across Google's Product Ecosystem

Veo technology already powers video features across multiple Google products, including YouTube Shorts, Google Photos, Google Vids, and the dedicated Flow filmmaking studio 3

3

. This broad integration demonstrates how generative AI video is becoming embedded infrastructure rather than standalone novelty. For developers, Veo 3.1 Lite opens the door to building similar features into their own apps without enterprise-level budgets.

Early testing of the model shows fast generation times, with 8-second videos rendering in under one minute 4

4

. Quality degradation compared to Veo 3.1 Fast appears minimal, making it suitable for content creation workflows where speed and volume matter more than cinematic precision. Google emphasized that this release represents just the beginning, teasing "more updates soon" as part of its commitment to making video generation more available to developers 3

3

.

What This Means for the AI Video Market

The contrast between Google's expansion and OpenAI's retreat highlights a fundamental reality: AI video generation remains compute-intensive and expensive to operate at scale 1

1

. While OpenAI struggled with the economics of running Sora as a consumer product, Google is betting on a developer-first model where costs distribute across thousands of applications. This approach could determine whether AI video becomes ubiquitous or remains confined to high-budget use cases. With pricing dropping across Google's entire lineup in one week, the barrier to entry for building video-enabled applications just got significantly lower.