Google has introduced a new entry-level AI video model called Veo 3.1 Lite, available to developers through the Gemini API. The model is priced at less than half the cost of Veo 3.1 Fast, the mid-tier option in Google’s video generation lineup. It supports both Text-to-Video and Image-to-Video generation in landscape and portrait formats, at 720p and 1080p resolution. Video length can be set at 4, 6, or 8 seconds, with pricing that scales accordingly.
The cost reduction is substantial when compared to earlier options in the same family. The original Veo 3.1 ran approximately $0.40 per second of generated video with audio, while Veo 3.1 Fast came in at $0.15 per second. Veo 3.1 Lite brings that figure down to $0.05 per second at 720p, a threshold that makes high-volume video applications more financially accessible for smaller developers and creators. Google has also announced that pricing for Veo 3.1 Fast will drop further on April 7, extending cost reductions across the entire Veo 3.1 model family.
Early testing of the model showed fast generation times with limited quality degradation. An 8-second video, the longest available duration, was produced in under one minute. Prompt adherence was described as respectable, with only a minor issue observed in on-screen lettering. The quality gap between Veo 3.1 Lite and Veo 3.1 Fast was noted as less pronounced than the gap between Veo 3.1 Fast and the original full version of Veo 3.1.
The pricing shift comes at a notable moment for the AI video industry. OpenAI‘s video product Sora was reportedly burning $15 million per day before the company announced it was shutting the product down entirely. OpenAI stated it is now pivoting to world simulation research to advance robotics. A reported $1 billion deal with Disney was also caught up in the fallout. The episode underscored how unsustainable costs have been a persistent challenge in AI video generation.
Google’s own history with video AI reflects a similar tension between capability and cost. Veo 3 launched in May 2025 as a flagship showcase, capable of generating not just video but full soundtracks including ambient noise, effects, and dialogue. Veo 3.1 followed in October, competing directly with Sora 2, but its pricing remained a barrier for developers looking to build at scale. The introduction of Veo 3.1 Lite is a direct response to that constraint.
Chinese competitors have been pressing on the economics of AI video for some time. Kuaishou‘s Kling AI has offered comparable video generation at prices lower than Google’s higher-tier plans. Tencent‘s Hunyuan Video went further by releasing an open-source model at no cost, timed to coincide with the attention surrounding OpenAI’s Sora launch in 2024. These competitors have demonstrated that the competitive battleground in AI video is not limited to output quality but extends to pricing strategy.
At the professional end of the market, tools such as Utopai‘s PAI are targeting a different segment, offering long-form cinematic storytelling with consistent characters, detailed storyboards, and scene-level AI editing. PAI is priced at $100 for 10,000 credits, signaling that serious creators are seeking greater control over the production process rather than simple generation. Veo 3.1 Lite is not positioned to compete directly in that space, nor is it aimed at undercutting the lowest-cost Chinese alternatives on price alone. Instead, it targets developers who need to integrate video features into applications at scale without incurring prohibitive API costs on each iteration.
Originally reported by Decrypt.
