Glossary
AI video glossary
The terms you'll run into when making videos with artificial intelligence, each explained in one sentence, without jargon.
- AI avatar
- A digital person who presents the video and moves their lips in time with the voice. It can be invented (with a visible AI notice) or your own avatar, created only with your consent.
- AI dubbing
- Translating a video into another language by redoing the narration with a voice in that language. Full dubbing also translates on-screen text, captions, title and description.
- AI video
- A video whose production (script, images, voice, music, editing and captions) is done by an artificial intelligence system from an instruction. The person brings the idea and the final approval.
- Auto-reframe
- Turning a landscape video into vertical by following the face of whoever is speaking in each shot, instead of blindly cropping the center.
- B-roll
- The supporting footage shown while the voice speaks: the product, the place, a detail. With AI it can be generated, found in licensed libraries or taken from your own photos.
- Batch videos
- Producing many videos at once in the same format, for example one per product in a store. On CreaVideosConIA it's up to 100 per batch, with the total price guaranteed before starting.
- C2PA (Content Credentials)
- An open standard that signs, inside the file, how a piece of content was made, for example «made with generative AI». Every CreaVideosConIA video comes out with that provenance signature.
- Clips (long video to shorts)
- Cutting a long video (a webinar, a class, a podcast) into several short videos with its best moments, reframed vertically and captioned.
- Credits
- The unit production is paid in: each scene spends according to the models it uses. For reference, the measured AI cost of a 60-second video is about US$0.21.
- Deepfake
- A video or audio that imitates a real person without their permission to make it look like they said or did something. CreaVideosConIA never uses real faces or voices without verified consent.
- Dynamic captions
- Captions that appear word by word in time with the voice and highlight the word being spoken. They make the video understandable without sound, which is how much social video is watched.
- Guaranteed quote
- A video's maximum price, shown and reserved before producing it. When it's done the real cost is charged, capped at the quote; on average customers pay 87% of the quote.
- Hook
- A video's first line or image, designed to stop the scroll. In short video it plays out in the first 2 seconds: a question, a surprising fact or a concrete promise.
- MCP (Model Context Protocol)
- An open protocol that lets AI assistants and agents use external tools. With CreaVideosConIA's MCP server, an agent can quote and produce videos for you.
- Open models (open weights)
- AI models whose weights are published and can run on your own servers. They're the default on CreaVideosConIA because they lower the cost; paid models remain an option.
- Prompt
- The text you use to ask an AI for something. For video, a good prompt states the topic, audience, tone and goal; for example: «3 mistakes when selling on Instagram, for small business owners, friendly tone».
- Retention
- The share of a video the audience watches before leaving. It's the metric that weighs most on TikTok, Reels and Shorts; CreaVideosConIA reads it from your accounts to learn which hooks and lengths work.
- Safe zones
- The areas of the screen not covered by each network's buttons, account name and caption. Text and captions must stay inside them; every network has its own.
- Script
- The text of what is said and shown in each scene. Reviewing it before producing is the cheapest way to fix a video: changing a sentence costs nothing, redoing a produced scene doesn't.
- Storyboard
- A video's sequence of scenes with what is seen, heard and how long each lasts. On CreaVideosConIA you see it with each scene's price before producing.
- Synthetic content
- Images, voice or video generated or altered with AI. TikTok, Instagram and YouTube ask you to label it when it's realistic; CreaVideosConIA labels it when publishing and marks generated people with a visible notice.
- Synthetic voice (TTS)
- A voice generated by AI from text (text-to-speech). Today's voices have natural pauses and intonation and exist in many languages and accents.
- Text-to-video
- Generating a video from text. It can mean a model that generates moving clips or, as on CreaVideosConIA, the whole pipeline: from one sentence to a finished video.
- Vertical video (9:16)
- The phone-screen format (1080 × 1920) used by TikTok, Instagram Reels and YouTube Shorts. The same video can also come out in 16:9 for YouTube and 1:1 or 4:5 for LinkedIn.
- Voice cloning
- Creating a synthetic voice that sounds like a real person from their recordings. On CreaVideosConIA it's only done with that person's consent, verified by voice, and the videos carry an AI notice.
Your next video is one idea away.
Create a free account, pick your industry and watch your first video being produced in about 4 minutes, with its cost to the cent.
No card · trial credits · ES/EN/PT