demobook

Google AI Studio: Cinematic generation with Google Veo 3.1

Demo summary

The user demonstrates Google Veo 3.1's 'Normal' high-quality mode to generate a cinematic scene with realistic audio.

Step-by-step

  1. Switch the model to Google Veo 3.1
  2. Navigate to the video mode setting
  3. Select Normal mode
  4. Set the resolution to the highest available
  5. Ensure version V3.1 is toggled on
  6. Click Create

Options

  • Fast mode for quicker generation

Watch out for

  • You must have Normal mode selected to achieve the highest quality output

Tips

  • Use Normal mode despite the name; it is significantly better than Fast mode
  • Keep the version on V3.1 as it is currently the highest available

Highlights

It is not normal by any means. It is much, much better than fast.

All demos from “Best AI Video Generators in 2026 (Most Realistic)

  1. 0:500:50Generate realistic video with Sora 2The creator demonstrates using Sora 2 within the OpenArt interface to generate a high-resolution, 12-second video from a complex text prompt.OpenArtText to Video
  2. 2:150:16Create vlog-style content with Sora 2A demonstration of Sora 2 generating a realistic 'vlog style' video of a winter day in New York City, showcasing environmental interpretation.OpenArtText to Video
  3. 2:520:35Image-to-video product ad with Sora 2The user uploads a static image of headphones and uses Sora 2 to generate a UGC-style video ad with a person demonstrating the product.OpenArtAI UGC Video Generator
  4. 4:270:18Text-to-video generation with Kling 2.6The video shows the selection of the Kling 2.6 model in OpenArt to generate a video with saturated colors and integrated audio.OpenArtText to Video
  5. 5:090:23Simulate cockpit environment with Kling 2.6A demonstration of Kling 2.6 generating a stylized video of a pilot in a thunderstorm, focusing on environmental effects and audio sync.OpenArtText to Video
  6. 5:460:18Animate static portraits with Kling 2.6The creator uses an image of a grandmother with cookies as a base for Kling 2.6 to animate a winking smile and camera interaction.OpenArtImage to Video
  7. 7:170:28Multi-shot camera control with Wan 2.6Demonstration of the multi-shot prompting feature in Wan 2.6, allowing for distinct camera angle changes within a single AI generation.OpenArtAI Video Generator
  8. 8:070:40Image-to-video scene creation with Wan 2.6The user demonstrates animating a retro-futuristic diner image by adding a UFO fly-by using Wan 2.6.OpenArtImage to Video
  9. 10:320:31Complex motion tracking with SeaDance 1.5 ProThe creator shows how SeaDance 1.5 Pro handles complex martial arts movements using skeletal tracking to maintain limb consistency.OpenArtAI Animation Generator
  10. 11:270:31Start-to-end frame interpolation with SeaDance 1.5 ProA demonstration of providing both a starting and ending image to SeaDance 1.5 Pro to control a specific dance sequence.OpenArtImage to Video
  11. 12:350:39Cinematic generation with Google Veo 3.1CurrentThe user demonstrates Google Veo 3.1's 'Normal' high-quality mode to generate a cinematic scene with realistic audio.Google AI StudioText to Video
  12. 13:380:25Dolly shot camera instructions in Google Veo 3.1A demonstration of using specific camera movement keywords like 'wide angle dolly in' to control the output of Google Veo 3.1.GeminiText to Video
  13. 14:230:36Room transition generation with Google Veo 3.1The creator uses a messy office start frame and a clean studio end frame to have Google Veo 3.1 generate a smooth room transformation transition.OpenArtImage to Video
  14. Watch “Best AI Video Generators in 2026 (Most Realistic)” →

Text to Video

  1. 12:350:39Cinematic generation with Google Veo 3.1CurrentThe user demonstrates Google Veo 3.1's 'Normal' high-quality mode to generate a cinematic scene with realistic audio.Youri van Hofwegen
  2. 7:121:46Create AI videos with Google AI Studio (Veo)The demo shows how to use the 'Generate Media' and 'Veo' options within Google AI Studio to turn text prompts into high-quality video clips.AI Playground
  3. 5:510:47Generate cinematic video with Google Veo 3.1The creator demonstrates the difference between a basic and a detailed prompt in Google Veo 3.1, showing a man answering a rotary phone in a gritty, neon-lit cinematic style with synchronized audio.AI Master
  4. 16:001:05Generate action scenes with Google Veo 3.1A demonstration of generating a high-action Michael Bay style clip featuring a cowboy riding a T-Rex through New York City, complete with explosions and synchronized sound effects.AI Master
  5. 1:420:35Generating a boxing match video with dialogueGoogle Veo 3 is shown generating a video of two boxers where one character delivers a knockout and speaks a specific line of dialogue with lip-syncing.@SocialtyPro
  6. 3:540:38Multi-character dialogue generationThe user shows how to prompt Google Veo 3 for a complex scene involving two men on a baseball field having a back-and-forth conversation with specific dialogue cues.@SocialtyPro
  7. 10:100:28Choreographed combat generation with Google Veo 3.1Google Veo 3.1 is shown generating a fight scene, demonstrating cinematic movement but lacking physical impact force.Youri van Hofwegen
  8. 10:160:25Generate cyberpunk video with Veo 3.1The user selects the Veo 3.1 model in the playground and generates a cinematic drone shot of a neon-lit cyberpunk city from a text description.AI Master
  9. 1:430:48Generate a dancing robot video with Veo 2A text prompt for a 'realistic looking robot dancing on the beach' is entered into Google AI Studio, and the resulting AI-generated video is played back.Moe Lueker
  10. 10:320:41Testing Google Veo 3 native audio and speech generationA series of generated clips from Google Veo 3 are shown to test if the model can correctly sync speech and audio for two characters at a bus stop. The results show various failures including garbled speech and incorrect character assignments.MisterQuigleyAI
  11. 11:220:08Evaluating character movement and speech in Google Veo 3A clip generated by Google Veo 3 shows a man in a red shirt asking for the time, demonstrating the model's ability to handle character movement alongside speech, despite some visual glitches.MisterQuigleyAI
  12. 3:210:30Test physics with breaking wine glassThe host tests the physics engines of both Veo 3.1 and Omni Flash by prompting a wine glass falling and breaking, showing the current limitations of both models.Creative AI Show