Are you drowning in the AI-video tools jungle? Me too!
But watch this. It might be the best 7 minutes of your week.
There’s a flood of AI video tools out there.
Which ones actually work? How to choose?
The Wall Street Journal did the hard work for us.
Here’s the exact stack they used to produce this short film:
▸ Midjourney, for character creation.
▸ Runway Act-One, to animate characters with real-life gestures.
▸ Runway References, to maintain visual consistency across scenes.
▸ Google Veo 2, for basic image animations.
▸ Google Veo 3, for enhanced animations and background sounds.
▸ Suno, for background music.
▸ ElevenLabs, for the robot's voice and the lady's voiceover.
▸ Any video editor to stitch it all together.
The result is impressive, but not jaw dropping.
Because it's made with consumer Ai tools only, hence not perfect.
But the point is that anybody can achieve similar results by using the exact same tools.
The reality is that many of the amazing Ai videos we see online are actually heavily edited by (semi)professional video makers using professional tools.
This is not the case and I love it.
They even share the final cost, around $1'000. Very realistic.
One element is still missing: lip-syncing.
Getting it right is still the hardest challenge. No wonder they didn't include it in this video.
But you can try Runway’s Lip-Sync tool, it gets fairly close.
Otherwise, Google Veo 3 can make characters talk in perfect lip-synch, but not with uploaded original audio.
Right now, the combo of Runway + Veo 2/3 might just be the most powerful AI video stack out there.
And remember,
this is the worst quality we’ll ever see. It only gets better from here.
