Guide

How to make an AI video in five steps

Pick the kind of AI video, write a script that fits the length, generate, edit, then export to each platform's specs. With the clip limits and per-second prices of the current models, so you can plan the shots.

Model limits and prices checked November 10, 2026

How to make an AI video in five steps
To make an AI video: (1) pick the type: avatar, text-to-video, motion graphics or generated clips; (2) write a script at about 150 words a minute (75 words for 30 seconds); (3) generate: most models make 4 to 15 seconds per clip; (4) edit the clips, voice and music together; (5) export a 16:9 master and a 9:16 cut.

Step 1: Which kind of AI video do you need?

"AI video" covers four quite different outputs. Pick the one that matches the job before choosing a tool.

TypeWhat you getTools and modelsGood for
A presenter talking to cameraAn AI avatar reads your scriptHeyGen, Synthesia, Colossyan, ElaiTraining, onboarding, updates, translated versions
A video from a script or articleStock or generated scenes with an AI voice and captionsinvideo AI, Fliki, Pictory, Lumen5, VislaYouTube explainers, social posts, blog-to-video
Generated clips from promptsRaw shots you edit together (no story, no voiceover)Veo 3.1, Seedance 2.5, Kling 4.0, MiniMax H3, Runway, Pika, LumaB-roll, ads, music videos, concept shots
Motion graphicsAnimated type, UI and shapes, designed scene by scenenod (from your product page), Hera, Jitter, MotionVidSoftware launch and explainer films, product ads

Prices and free plans for each tool are in best AI video generators by use case; model specs in AI video models compared. Explaining a product? See AI explainer video generators and AI motion graphics tools.

Step 2: How do you write the script?

Write the voiceover first, then the on-screen text and shots for each line. Spoken scripts run at about 150 words a minute, so the length sets your word budget:

Video lengthWords of voiceover (at 150 wpm)Typical use
15 secondsabout 38Ad, teaser, Story
30 secondsabout 75Social ad, product teaser
60 secondsabout 150Explainer, feature launch
90 secondsabout 225Launch video, walkthrough

Step 3: How do you generate the video?

Avatar and text-to-video tools take the whole script and build the video; you pick a voice, avatar or template and regenerate scenes you don't like. Model clips are generated one shot at a time, so a 30-second video is several generations:

ModelMax per clipClips for 30 sRate (tier)30 s, one take
Pika H3 Flex15 s2$0.033/s (H3 Flex 768p, Starter monthly)$1.00
Gemini Omni Flash15 s2$0.07/s (720p)$2.10
MiniMax H315 s2$0.08/s (MiniMax-H3, 768p)$2.40
Veo 3.18 s4$0.10/s (Veo 3.1 Fast, 720p, with audio)$3.00
Wan 3.030 s1$0.10/s (720p)$3.00
Runway Gen-4.510 s3$0.12/s (Gen-4.5)$3.60
Kling 4.030 s1$0.126/s (Runway API, 720p (to Nov 30))$3.77
Seedance 2.530 s1$0.231/s (720p)$6.93

Computed from each developer's published per-second price (Kling 4.0 via Runway's API; Pika from its Starter credits), checked November 10, 2026. Luma prices its API in credits without a dollar rate. Budget for several takes per shot.

Step 4: How do you edit an AI video?

Put the clips on a timeline in script order, trim each to the line it covers, then add the voiceover, music and captions. Runway Gen-4.5 and Luma Ray3.2 generate no sound, so add it all here; models with native audio still need levelling against your voiceover. Make it readable with the sound off: burn in captions or on-screen text.

Editors that help: Descript (edit by transcript), Kapwing and VEED (browser, auto-subtitles), Runway's Aleph 2.0 (change what's in a shot). Compared in best AI video editing apps.

Step 5: How do you export for each platform?

Export one 16:9 master at 1920×1080 for YouTube, LinkedIn, X and your site, and a 9:16 cut at 1080×1920 for Reels, Shorts and Stories. Each platform's length, size and ratio limits are on social media video sizes; check a finished file with the video specs checker (nothing uploads), and plan file size with the video file size calculator.

How to make an AI video: FAQ

How do I make an AI video for free?
Use a tool with a free plan and accept its limits: Colossyan (20 minutes a month), Synthesia (10 minutes), Lumen5 (5 videos of up to 2 minutes, 720p) or HeyGen (3 videos up to 1 minute), all with a watermark or outro by their own pricing pages. Runway's 125 one-time credits make about 10 seconds of Gen-4.5. Full list: free AI video generators.
How do I turn a script into an AI video?
Paste the whole script into a text-to-video tool (invideo AI, Fliki, Pictory, Lumen5 or Visla) or an avatar tool (HeyGen, Synthesia, Colossyan, Elai): it splits the script into scenes, adds stock or generated visuals, an AI voice and captions, and you regenerate the scenes you don't like. With a model such as Veo 3.1 or Kling 4.0 you turn each line into a shot prompt instead and edit the clips together (steps 2 to 4 above). Free plans: free AI video generators.
How long does an AI video clip last?
Per generation: 4 to 8 seconds for Veo 3.1, 3 to 15 for Gemini Omni Flash and MiniMax H3, 2 to 10 for Runway Gen-4.5, 5 or 10 for Luma Ray3.2, and up to 30 for Seedance 2.5, Kling 4.0 and Wan 3.0, by each developer's docs. Longer videos are several clips edited together, or extensions.
How much does it cost to make an AI video?
For 30 seconds of generated footage at about 720p, one take costs from $1.00 (Pika H3 Flex) to $6.93 (Seedance 2.5) at published API rates. Subscription tools charge $8 to $30 a month for entry plans. Retakes multiply the cost.
Can I still use Sora to make AI videos?
No. OpenAI's API docs say the Sora 2 models and the Videos API were shut down on September 24, 2026. See Sora 2 for the models people use instead.
How long should an AI video be?
As long as the platform and the message need: 15 to 30 seconds for an ad or teaser, 60 to 90 seconds for an explainer or launch video. In our study of 150 top Product Hunt launches the median launch video ran 78 seconds.

More: AI video models · Free AI video generators · Alternatives · Compare

Making a launch video?

Paste your product page link. nod does all five steps and hands you the export.