Exploring Google VEO 3.1

Exploring Google VEO 3.1 + Elevenlabs

Project Context

Architectural animations traditionally take a while to render. I remember sitting in the computer lab at the Southern California Institute Of Architecture curating scenes in Maya with the object to be rendered, adding lighting, texture etc and waiting for frames by morning which I would then stitch using an Adobe product to create a fly through. Wow! That's a long process that thought me allot about how to animate.


Fast forward to the year 2026, we have Google VEO 3.1 and I wanted to test out its capabilities with the following experiments and a goal of hopefully stitching multiple videos. I used my SCIArc thesis as a test subject as its my creation.


I have always used Imovie to create videos, and decided to test out Elevenlabs this time which was very user friendly. The following was the outcome.

Exp 1 : Using a 3D chunk and sectional render

I started by combining two renders from my thesis and added my thesis description. The following was the result. I used the Video setting with a start and end frame

Exp 2 : Using just a sectional render

Next I wanted to make the sectional render come to life with people walking inside.

Exp 3 : Bringing the white silhouttes to life

The LLM just used the silhouttes and I wanted to bring the people to life. I asked it to do that, the out put was interesting however I cant seem to know what accent it is!

Exp 4 : Animating the 3D chunk

Next I tried creating a 360 fly through of the 3D chunk, I liked this iteration!

Exp 5 : Animating the 3D chunk with inhabitants

Next i tried to add people to the chunk and had to use a second prompt to get it close. Alas, I used all my video generation credits so thats the end of my experiment for now!

Reflections

Overall I think VEO 3.1 is a great model to generate video. I did run into hallucinations but they were minimal. I noticed that the LLM did take the liberty of adding voiceovers which was a nice touch but I wish I could control the voiceovers more!


I decided to use Elevenlabs' studio to stitch together the video and voice overs which was really smooth! Overall a fun project.

Interested in working together? Let’s collaborate!

Interested in working together? Let’s collaborate!

Interested in working together? Let’s collaborate!

© Tarun hari 2026

The Spatial Organization Orchestrator