Mindket
English
English简体中文
Sign inSign up
StoreMCPSkillsCalculatorLearnPriceBlog
Discover
Research308Writing59Business95Legal13Finance & accounting33Marketing & SEO172Developer1,379Productivity382Education81Companions14
Help and support
Language
English简体中文
IntermediateFree

AI video: Sora, Veo, video prompts and open models

Mindket editors · Updated · 4 lessons · 31 min

A black banner titled "AI video" over the line "Sora, Veo, video prompts and open models".

This free course teaches AI video in four lessons. Sora is covered as closed, with the prompting that still works. Then come a first clip with Veo in the Gemini API, how to write a video prompt, and how to run open models such as Wan2.2 on your own computer. All of it follows the vendors' and projects' own pages.

What you'll be able to do

  • Explain what ended with Sora, and reuse OpenAI's shot template in any video tool.
  • Make a first clip with Veo in the Gemini API, from a prompt or from a start picture.
  • Write one shot as a video prompt, with the camera, the light and the sound, and keep a character the same.
  • Pick an open video model that fits your video memory and license, then run Wan2.2 or use ComfyUI.
All courses

Who it's for

For people who already make images or use a chatbot and want to make short videos. You need no coding to read the lessons, but lesson 2 and the Wan2.2 steps in lesson 4 use a terminal, and running an open model needs a graphics card with enough memory.

Lessons

  1. 1Sora tutorial: what ended, and the prompting that still worksSora's API shut down on September 24, 2026. Keep what still works: OpenAI's shot template, its settings and where each idea lives now.8 min
  2. 2Veo tutorial: make an AI video with the Gemini APIA Veo tutorial for the Gemini API: install the SDK, make a first clip, set a start and end frame, add reference images, extend the clip and meet Gemini Omni.7 min
  3. 3AI video prompts: how to write a shot, add sound and keep a character the sameHow to write AI video prompts that work: one shot, one action, a camera, light, sound, timing and a character who stays the same, from the vendors' own guides.8 min
  4. 4Open AI video models: run Wan, LTX, HunyuanVideo or CogVideoX on your own computerRun open AI video models on your own computer: Wan2.2, LTX-2.5, HunyuanVideo-1.5 and CogVideoX, with the memory, license and first command each project states.8 min

Frequently asked questions

Is this AI video course free?

Yes. Every lesson is free to read. The Veo lesson uses a paid Gemini API feature and the open models need a graphics card, so each lesson says what its tools ask for.

Can I still use Sora?

No. OpenAI removed the Sora 2 models and the Videos API on September 24, 2026, and lists no replacement. Lesson 1 keeps what still works, which is its prompting guide.

Do I need to write code?

Not to read the lessons, but two of them use a terminal. The Veo lesson calls the Gemini API from Python, and the open models lesson runs the Wan2.2 commands, with ComfyUI as a route that skips them.

Where should I start?

Lesson 3 on prompts works with every tool, so start there if you already have a video tool. Start with lesson 2 to make a first clip with Veo, or lesson 4 to run a model on your own computer.

What should I learn after this course?

AI voice and music is next. It starts with text to speech and goes on to voice cloning, songs and dubbing, the sound a silent clip still needs.

Share this course

Make a first clip

Lesson 2 walks through a first Veo clip, from the API key to the saved MP4.

Open the Veo lesson

All courses · All articles