AI tools
December 4, 2023

Pika AI Can Generate Videos With Simple Text Prompts

Pika is the latest text-to-video tool to enter the AI video party.

Jim Clyde Monge
by 
Jim Clyde Monge

AI models that can generate videos from text or images have woefully lagged behind conversational and image generator AI tools for far too long.

But the text-to-video space is finally catching up.

Just last week, two huge startups in the AI sphere, Stability AI and Meta AI, unveiled their own text-to-video tools—Stable Video Diffusion and Emu Video. Both seem promising, but in my view, they still have progress to make.

Today, another startup, Pika Labs, is about to join the party with their own AI video generator tool called Pika 1.0.

What is Pika?

Pika is a platform that can generate videos from text or images using artificial intelligence. The aim of the tool is to revolutionize how videos are made and edited through a state-of-the-art video foundational model.

a crowboy riding a horse on the desert. Image from Pika labs
Image from Pika labs

According to its creators, anyone can channel their inner Scorsese or Spielberg.

Our vision for Pika is to enable everyone to be the director of their own stories and to bring out the creator in each of us.

I couldn’t agree more. Their goal resonates deeply with my belief that these AI technologies could empower ordinary people like you and me.

Here’s how it works

I tried Pika on their Discord server. Under one of the “generate” channels, hit the “/” key on your keyboard to bring up the menu.

Pika labs discord channel
Image by Jim Clyde Monge

From here, you can either generate a video from an image or a text description. Let’s try an example with this prompt:

Prompt: An astronaut riding a horse on the moon

It took less than a minute for the AI to generate a 3-second video in the same channel on Discord.

Pika labs discord channel
Image by Jim Clyde Monge

Here’s what the final results look like:

A video of an astronaut riding a horse in space
GIF by Jim Clyde Monge

How cool is that? I think it looks awesome. The astronaut and horse are small in the background, though. I wish they were more prominent, but it’s still an impressive start.

Here’s another example that takes an image as an input. This is the image input generated in Midjourney:

Image by Jim Clyde Monge

This is the output video from Pika:

A video of a crow flying through an eerie wasteland. Generated by PIKA labs
GIF by Jim Clyde Monge

What do you think?

Again, Pika animated the image nicely. But the video lacks dynamics, in my opinion.

Pika 1.0 will include the following features:

  • Text-to-Video and Image-to-Video
  • Video-to-Video: Transform your existing videos into different styles, including different characters and objects
  • Expand: Expand the canvas or aspect ratio of a video
  • Change: Edit video content with AI, such as changing someone’s clothing, adding in another character, or changing the environment
  • Extend: Extend the length of an existing video

How to access Pika?

Currently, Pika is accessible in two ways:

If you want to know more details and updates on Pika, check out their X and Product Hunt pages.

What’s possible

If Pika lives up to its potential, high-quality video creation could become much more accessible. This could greatly empower amateur filmmakers and creators.

Imagine how this could empower a new breed of filmmakers and storytellers. This could spark a revolution in the film industry.

And therein lies the opportunity, because regular folks like you and me, with the right prompt and input image, can achieve content that we could only have dreamt of months ago.

Aren’t you excited to produce your own film?

Final Thoughts

Overall, I am very happy to see that the AI video space is finally heating up. Big companies like Google and Microsoft are surely working on their own versions of the same tool too, and I can’t wait to see them unveiled soon.

While still early, Pika hints at the future possibilities. I’m eager to see how it develops and helps unlock more creativity.