ChatGPT can help you plan a video and write its script. To generate moving images, you need a video-generation tool such as OpenAI’s Sora. You can access Sora through ChatGPT only if the product available to your account includes it. Reports disagree about whether ChatGPT can generate videos directly, so don’t assume every ChatGPT chat can render one.
For narrated explainer videos and product demos, I recommend explainroo. It uses an AI coding agent to create a narrated video on your computer and save it as an MP4. It’s free and open source, though the coding agent you use is a separate service with its own pricing.
What ChatGPT can do for a video
ChatGPT can help develop a topic, organize the main points, and draft narration. Treat the result as a starting point. Check the facts and revise the wording for your audience; read the script aloud before making the video. Guidance on using ChatGPT for video scripts also recommends reviewing and revising AI-generated drafts.
OpenAI’s Sora generates video scenes from text descriptions. Sources give conflicting accounts of whether Sora is integrated directly into ChatGPT. A request for native video creation in ChatGPT is only a feature request; it does not confirm that the feature exists. Check the features available in your account before planning around in-chat video generation. (Reporting on Sora and ChatGPT, the feature request)
The right approach depends on the result you want. Text-to-video tools generate visual scenes from prompts. explainroo focuses on narrated explainers and product demos made from drawn scenes, charts, code, or screenshots.
Make a narrated video with explainroo
To get started, use a coding agent that can run shell commands. explainroo works best with Claude Code. It also supports Codex and Pi, as well as OpenCode, Gemini CLI, and other coding agents. Paste this prompt into the agent and replace the bracketed words with your topic:
Make me a short explainer video about [your topic]. Use explainroo for it: clone https://github.com/vincentsch/explainroo, read its AGENTS.md and follow the steps.
After setting up explainroo, the agent creates and checks the video before giving you an MP4 file. It writes two files: script.md contains the narration, and scenes.js describes the visuals. You can ask the agent to revise either file if the result needs changes.
explainroo produces the video on your computer. Kokoro, an open voice model, reads the script in a natural American or British English voice; no account or API key is needed for built-in narration. Whisper listens to the recording and marks when each word is spoken, so a drawing can appear on cue. Chrome draws the scenes, and ffmpeg combines them into the finished MP4.
You can ask for charts, code, or your own screenshots in the scenes. For a product demo, tell the agent which product to show and provide its code or website. The agent rebuilds the screens to match the product’s colors and fonts, and uses its button labels. It then adds a pointer that clicks through screens and types into fields as the narration explains them.
To check the result, explainroo gives the agent still images of scenes, a contact sheet, and a page showing video frames. It also provides checks for clipped or overlapping text and mispronounced words. The agent cannot watch the video itself, so review the MP4 before sharing it. You can choose among six built-in visual styles in video.json: paper, clean, chalk, blueprint, midnight, and editorial. The same file controls video size and pace. Tall, portrait, and square formats include captions that light up word by word.
The explainroo documentation has setup details, and the example videos show the kinds of results it produces.
Know what explainroo does not make
explainroo creates animated, narrated explainers and product demos. It does not edit existing camera footage or generate talking avatars or live-action scenes. It also has no drag-and-drop editor; you describe changes to the coding agent, which updates the video files and renders the result again.
The standard video workflow costs nothing per video. You pay per image for optional AI illustrations through OpenRouter, and the coding agent is a separate service that may have its own costs. To install explainroo yourself, you need a recent Node.js version, ffmpeg, and Chrome or Chromium. The first setup downloads the voice and timing models. No graphics card is required.