TL;DR Using Google’s new Veo 3.1 video model, we created a breathtaking 1 minute 40 second FPV drone flight through mountain valleys, and it took just 15 minutes to generate.
Imagine soaring through alpine valleys, gliding between snowy peaks, and diving toward rivers that twist like silver ribbons below, all without leaving your desk. That’s exactly what we did using Veo 3.1, Google’s latest generative AI video model.
Our test video, a first-person drone flight across a mountain range, captures the freedom and exhilaration of flying while showcasing just how far AI video tools have come. The result is stunningly realistic: sunlight glinting off ridges, smooth motion through tight turns, and even the soft rush of wind that accompanies the journey.
What’s most incredible? The 1-minute 40-second clip took only about 15 minutes to generate … proof that we’ve entered an era where high-quality aerial cinematography is within everyone’s reach.
Veo 3.1 is Google’s upgraded AI video generation model, designed for more realistic, longer, and higher-fidelity results. It supports single clips up to one minute, full HD (1080p) resolution, and synchronized, natural-sounding audio. Veo’s improved prompt adherence and detail control make it ideal for cinematic content, professional storytelling, and creative exploration.
Key upgrades include:
First & last frame guidance for precise scene transitions
Scene extension for longer, coherent storytelling
Horizontal & vertical aspect ratios for any platform
Enhanced realism and sound design that bring visuals to life
Veo 3.1 moves beyond traditional text-to-video generation. It feels like directing a film rather than typing a command.
Veo’s new Flow-based SceneBuilder lets users grow or modify videos with continuity in mind. You can extend a clip’s final frame into new terrain, add cinematic transitions, or adjust lighting and style between sequences, all without breaking immersion.
In our FPV drone project, SceneBuilder allowed the AI to “keep flying” beyond the initial one-minute limit. By extending the final frame seamlessly, Veo 3.1 stitched together multiple generative passes into one continuous flight through valleys and canyons, a feat that would’ve required hours of manual editing before.
It’s like having an AI co-pilot who knows exactly how to maintain altitude, momentum, and atmosphere.
Another standout feature, Frames to Video, transforms any pair of images into an animated sequence, an invaluable tool for creative transitions. By defining a start and end frame, Veo generates motion between them, enabling smooth transformations or time-lapse-like effects.
This is perfect for creative storytelling. For instance, transforming a static mountain photograph into a sweeping drone ascent, or blending two perspectives into a single cinematic moment.
Veo 3.1 represents a significant step toward democratizing filmmaking. What once required professional drones, pilots, and post-production teams can now be achieved in minutes by anyone with imagination. Artists can storyboard worlds. Educators can visualize concepts. Filmmakers can pre-visualize entire scenes with photorealistic accuracy.
For us, this drone-through-the-mountains video wasn’t just a test. It was a glimpse of the future of creative storytelling, where AI turns imagination into motion.
In short: with Veo 3.1, anyone can fly their own drone, no propellers required.
Most current vision-language-action (VLA) models—such as OpenVLA, π0, RT-2, and RDT-1B—are “monolithic.” This means they…
Australian teams working with OpenAI models can now access the latest OpenAI models through Amazon…
AI models have clearly proven their ability to discover and exploit vulnerabilities without much, if…
The company is reducing pressure on workers to use artificial intelligence tools while encouraging them…
Self-driving cars are often controlled by deep learning models that sometimes fail in unexpected situations.…
Today, we’re excited to announce the availability of Claude Fable 5.1 on Amazon Bedrock and…