AI Short Video Generator Powered by OpenAI

Want to create a video with just a title?

You type a word, and in 60 seconds or less you have a finished video.

Without taking up your valuable time, and no need for a video editing program, you have your finished video. The system creates everything for you: writing the story, visual content, creating voiceovers and applying edits.

The system used to create the video uses OpenAI technology, with GPT-4 generating the text, DALL-E 3 creating the visuals and then Whisper API providing the sound and subtitle creation.

You can directly upload your new video to TikTok, Instagram Reels or Youtube Shorts.

The best part of this product is it allows you to save time. Typically creating one video requires hours of work including text, images, editing and sound. AI has compressed those hours of work into a few minutes. Producing content has gone from 20-30 times faster than before.

If you are someone that is always on the go, you can now easily automate this process and save yourself some major headaches.

There is no question, automation is the greatest advantage of the AI Video Generator.

No more working for free, everything can now be done with the push of a button!!

• Creating 20 videos an hour is a level of speed and efficiency that was previously unimaginable.
• Flexibility — you can request any type or style of video simply by providing a new request.
• Cost effective — an individual utilizing AI can perform the tasks of a team consisting of a script writer, designer, and editor.
• The generator automates the majority of the time consuming tasks associated with creating videos — you no longer need to search through stock libraries, write text manually, or spend time in the edit suite. Just choose a topic, set your parameters, and the system does everything else.

How does it work?

OpenAI technology provides the underlying capabilities for this service. There are three basic components to the system:

The GPT-4 engine produces a script. To generate a video, you input a request such as Create a Video on DeFi for Beginners, and the engine generates an outline with estimated timings and key messages for each section.

DALL-E 3 generates images for each of the three to seven scenes of the video.

Whisper API creates the voiceover in more than 50 languages and adjusts the tone of the voice for the intended audience, creates subtitles in synchronisation with the voiceover audio track.

There is an API infrastructure that links all three elements, providing a mechanism to gather all the pieces of a video and render/export them in the desired format. A typical 30-second video can be created in about 45 to 90 seconds. The entire process can be started with one request without any actions in between.

Step-by-step Process for Video Creation Using AI

Below is an outline of what you can expect from the video creation process:

Step 1 — Task Definition

In this step, you outline your idea and the technical details of the video including topic, length, type format (style), and distribute location (platform). Our system takes this input, compiles it into a command for our AI program(s) and generates a video from the start to finish.

Step 2 — Script Generation

The AI then uses GPT-4 to produce the video script, broken down into scenes and includes visual descriptions as well as a timeline.

Scene 1 (0:00 - 0:07):
Voiceover: "DeFi protocols processed 2.1 trillion dollars in transactions in 2024."
Visuals: Neon graphic depicting the growth of transaction volume.

Step 3 — Visual Creation

Our AI will also produce illustrations based on each script scene. You have the option to add animations (such as a parallax effect) to enhance your images. If you need additional stock video footage, we have built-in libraries for you to use.

Step 4 — Voiceover and Subtitles

Utilizing Whisper AI, we will create a voiceover for your video based on the user-created script and provide synchronized subtitles for the completed video. Timing accuracy is +/- 0.1 seconds.

Step 5 — Final Compiled Video File

The final step is to create an all-in-one video file that you can save to your PC, or post automatically on your social media through a variety of integrations. Depending on your subscription, you can run multiple video creations at once.

Integration and Technical Aspects of the AI Video Generator

Formats Supported

The AI video generator currently supports all of the major video formats:

• MP4 (H.264) — This format is widely used for social media and has a variety of aspect ratios, i.e., 9:16 (vertical) and 16:9 (YouTube). The recommended MP4 bitrate for HD video is 5-10 Mbps.
• MOV (H.265) — This is the best video format for high-quality editing, and it supports transparency in your video files. However, MOV files tend to be larger than MP4.
• WebM (VP9) — The perfect format for web use because of its low bit rate and good quality.

You can set export settings using either API or web interface, so you can select the resolution, format, frame rate, etc.

For more exotic formats such as GIF and MPEG-2, you will need to use another service to convert them to these formats.

Frequently Asked Questions (FAQ)

Can I use my own voice for the voiceover?

Yes, in Pro plans and higher, you can upload your own audio files, and the system syncs the voiceover with video as well as creates subtitles through Whisper.

Is the Russian language supported?

Yes, it is fully supported. GPT-4 creates scripts in Russian, Whisper creates voiceovers using a natural-sounding intonation accent, and DALL-E processes requests in Cyrillic. Subtitles are generated without any issues.

How long will the completed video be stored?

You can keep the completed video for 30 days if you are in the Starter plan. There are no storage limits for Pro plans and above; however, you may download the video right after it is created.

Can I modify the script after it has been created?

Yes, you can edit the script before the render starts. Once the render has been initiated, you must create a new session with an updated prompt. You can also export the script and the visuals for manual editing purposes.

What topics are not allowed?

You may not post content related to political campaigning, 18+ material, inciting violence, or using a third-party trademark (unless you have a license). If you violate any of these regulations, your account will be suspended without a refund. The full list of prohibited content can be found in the user agreement.

Can the generator be integrated into a Telegram bot?

Yes. The API reference includes an example of a Python-based Telegram bot that takes prompts, sends requests, and returns the finished video. The time to create a video is around 1 minute.

Reviews and Case Studies

Case 1: Cryptocurrency Education Channel

I used to spend 4 to 5 hours per video: writing scripts, finding images, editing, and recording my voice. Now with the generator, I can do everything in 10 minutes. I produce 3 to 4 videos daily and have increased my reach sixfold in just 2 months. The system automatically selects images for me, I no longer have to search for charts.

• Production speeds increased 24x.
• The cost of each video dropped from $50 to $0.99.
• CPM decreased 40% due to higher frequency of publications.

Case 2: Marketing Agency

Before the advent of the growth agency, the firm could not provide clients with video advertising due to costs. Now with the generator, 1 specialist can produce 10 videos per client each month. Revenue has increased 30%.

Clients are amazed at our turnaround time and quality. Traffic and inquiries from Reels and other sources are climbing. This represents a competitive advantage for us.

Case 3: Personal Brand of a Fitness Trainer

Before switching to the generator with API, trainer Dmitry would rarely publish videos because he struggled to find the time to edit them. Now, he creates 7 videos each week from voice notes by having them automatically transcribed to scripts and videos. He has increased his number of consultations by 4x.