Upscale an AI video without exposing that it is AI
Topaz Starlight Precise 2.6 and FLUX Video Upscale: which to try first, settings, list prices, and before-and-after video from our own upscale.
Notes on making AI video
What things cost, what breaks, and what we can actually measure. Prices and findings carry the date they were taken.
Guides
Start here
Practical guides to making AI video, built from our own takes, invoices and logs.
- AI video generators: how they work and how to get a finished film
- Best AI video generators in 2026, and what free really gets you
- Image to video AI: how to animate a photo, with real inputs and prompts
- How much does a 30-second AI video cost? Estimator and real numbers
- Kling vs Seedance vs Wan vs Veo: which AI video model is better?
- Consistent AI video: same characters, locations and light across shots
- AI video ads and UGC without filming: a production workflow
Featured collection
AI Video Price & Quality Index
The AI-video market changes every week. We are tracking it: dated prices now, comparable quality tests as they are run.
Quality scoring is not published yet. A rubric has to be defined and applied to more than one test round before any ranking would mean anything.
ComparisonsDated prices and like-for-like tests.
GuidesHow to direct a shot and evaluate the result.
Product notesWorkflow decisions, measurements and changes.
AI video prices, November 2026: Wan 3.0 vs Kling 4.0 vs Seedance 2.5
What 5 seconds of Wan 3.0, Kling 4.0, Seedance 2.5 and other models cost at list price on each platform, why Wan is cheaper for new shots and Kling for edits, and when a monthly plan beats paying as you go.
Best AI video generators in 2026, and what free really gets you
Picks by job from our own runs, and a dated table of what each free plan includes. Runway, Kling, Veo, Luma, Pika, Firefly, HeyGen, Synthesia, Canva, CapCut and InVideo.
Kling vs Seedance vs Wan vs Veo: which AI video model is better?
Kling 4.0, Seedance 2.5, Wan 3.0 and Veo 3.1 compared on price, specs, audio and what happened in our own takes, with the failures left in.
How much does a 30-second AI video cost? Estimator and real numbers
Raw generation, retakes, sound, upscaling and human time, separated, with real numbers from a 30-second film and an interactive estimator.
AI video with dialogue: which models do speech and lip sync
Which AI video models generate speech and lip sync in the same pass, what went wrong in our tests, and when to use a voice tool or a lip-sync pass.
AI video ads and UGC without filming: a production workflow
How AI video ads and UGC-style clips get made without filming: production steps, deliverables, real examples, limits and disclosure rules.
Video to video AI: replace people and places and keep the motion
Recasting people, locations and colours in a finished clip: four replacement routes judged on whether the people still move, why Wan 3.0 needs short runs, and why outpainting 9:16 to 16:9 shows.
How long does AI video take, and how to build longer films from scenes
Measured generation times from 69 of our own runs, the length limit per generation, and five ways to connect scenes into a longer film.
Camera moves for short AI B-roll: one small move, not a journey
Seven takes from one reel in MiniMax H3 and Google Flow: a camera locked to the subject, a crane to top-down in two seconds, how far small moves really went, the four-beat flight, and a camera skill to download.
Image to video AI: how to animate a photo, with real inputs and prompts
How to turn a photo into a realistic AI video, with real input pictures, the videos they produced, the exact prompts, model settings and costs.
AI video generators: how they work and how to get a finished film
What an AI video generator does, how text, image and reference generation differ, and the working route from an idea to a finished film. The place to start.
One AI video shot, seven routes: Google Flow, Seedance 2.5, Wan 3.0 and MiniMax H3
One miniature comedy scene shows how model choice and specialised prompt skills affect picture consistency, dialogue, camera moves and production costs.
AI video editing: prompts, video to video, object and background removal
Which AI tool to use for each editing job: prompt edits, video to video, background and object removal, recasting and restyling, with costs from our own runs.
Change the background and outfit in a talking-head video: seven AI routes
One selfie clip, seven ways to change the room and the clothes around the speaker: restyle, cut-out and photo-avatar routes, with the process, prompts and a reusable template.
Consistent AI video: same characters, locations and light across shots
How to keep the same character, location, lighting and camera angle across AI video shots: reference sheets, empty location frames, shared frames, and why fewer references work better.
Remove or replace an object in a video with AI: six tools compared
Seedance 2.5 Video Edit, Beeble SwitchX, Gemini Omni Flash, Higgsfield Genjutsu, Runway Aleph 2.0 and Kling O1 Edit for changing one thing in a finished clip, with a real Genjutsu job and the objects it missed.
AI video reference frames: how the start image, prompt and model shape a shot
Six running-commercial takes show how the reference frame, prompt direction and model choice combine to shape the final ad.
The flat rate that wasn’t: $0.134 advertised, $4.75 billed
A published “flat” per-second rate excluded the resolution we actually wanted. Reconciliation against one paid run exposed the 4.4× gap.
AI video cost per second: 16 options compared, September 2026
Sixteen service and model combinations reduced to a per-second rate, with the resolution, pricing basis and limits that make the headline numbers unlike each other.
