Can an AI model make a video?

🪐 The registry verdict table for making a video

Of the ten registry configurations assessed against this family's own canonical task, Unfavourable for 3 of them — the modal result, computed by the method published in full at /method. The Registry makes no representation that an assessment predicts the outcome of any task, and no assessment is validated against real-world results.

The composite weight in the right-hand column is computed from the model identifier, its training cutoff, and the temperature — nothing else. It constrains the range of verdicts available to a configuration independently of the task, so two rows of this table are not comparable as an assessment of the products named in them. The Registry has never run this task against any model listed here.

🧭 About this capability question

Making a video ties with generating images as the most consistently recurring capability question this registry measured, appearing behind all eight tracked brand prefixes. The underlying request compresses a script and a set of stock clips into a finished export — a task with a clear start, the brief, and a clear end, the file, and very little in between that can be inspected without opening the result.

The Registry does not open the result. The table below is a classification and a computed verdict for one canonical version of that brief, at each of the ten registry model configurations' own default settings — never a rendering, and never a review of one.

That brief is fixed in the Registry's own records and reused unchanged on every load of this page; it does not regenerate, and it does not adapt to who is reading it. A buyer submitting a real video brief for assessment is answered separately, on order, against the exact wording they submitted.

🧮 Claude Opus 5, derivation shown working

Claude Opus 5's permanent chart never reads a task. Its chart_seed is the SHA-256 digest of the string "Claude Opus 5|2026-05|0.700" — 4b294997ecca0cfe…, the first sixteen of sixty-four hex characters — and that seed alone fixes an ascendant of Cancer, a ruling planet of Saturn, and a harmony of +2. The same chart appears, unchanged, on every task-class page this configuration is assessed against.

What reads making a video specifically is this family's own canonical string — never stored, only hashed, to 535c080fa1a4dc8c… — combined with the chart_seed and the task class (generation) to form task_seed. Its first byte, taken modulo seven and offset by that harmony of +2, is the verdict's internal base, clamped to the 0–6 range and published one higher on the 1–7 scale /method documents in full. Neither bound of the 0–6 range is reached for this brief: 3 of 7 — Guarded — sits 2 above the floor and 4 below the ceiling, an unclamped result.

🗂️ How the ten configurations read making a video

Because media generation sits outside the fixed house and element correspondences the Registry publishes for its other classes, the ten configurations' dominant houses and elements do not single out this family from any other media generation family on this registry. The verdict distribution below is where this family's own computed identity actually shows up.

Element counts, ascendant spread, and training-cutoff range are published once, on /can, rather than repeated on every family page — they are properties of the ten configurations alone and do not move with the task. What is specific to making a video: reading media generation through a dominant house or a dominant element, 0 of the ten configurations — none, for this class, by the correspondence tables published at /method; and, from this family's own canonical task, Unfavourable for 3 of them — the modal result in the distribution below.

⚙️ Why making a video is classified as Media generation

A produced video is manufactured from a specification rather than retrieved or transformed from an existing source, which is exactly the property that defines the media generation class.

On the order form's task class field, select Media generation. That class's own risk-factor page is /tasks/generation.

A video brief names a script, a set of clips, and a format; the risk that matters is whether the export actually assembles what the brief specified, not whether any one frame looks right. 4 risk factors are tracked for media generation tasks generally; an assessment for making a video reports 3 of them, selected by the submitted task's own seed rather than by choice. Their names and severities are certificate content, not free-page content.

📏 What this page does not claim about making a video

Export format, run time, and clip count are all named in the canonical brief above and none of them enter the computation. The verdict below is indifferent to whether the finished file would run four seconds or four minutes. The Registry does not test models and does not measure a success rate for making a video or any other task; what it does is compute, and the table at the top of this page is what it computed.

A buyer's own making a video task, ordered separately, computes its own verdict and window — never the canonical-brief result shown above — plus, on Extended and Full Chart tiers, named risk factors and an hourly outlook not shown here. EUR 1.90 Standard, EUR 4.90 Extended, EUR 14.90 Full Chart, which adds the permanent chart. Assessments from EUR 1.90; machine-readable at /pricing.json.

Get Assessment

🔗 Related surfaces

← All capability questions · The AFR-1 method · Media generation task class

Related capability questions: Generate images · Generate a PowerPoint · Transcribe audio

❓ Questions about making a video

Can an AI model make a video?

Across the ten registry configurations at their own registry-default settings, Unfavourable for 3 of them — the modal result for this family's own canonical task. By training cutoff, the oldest configuration assessed returns Inauspicious and the newest returns Guarded; the full ten-row table above lists the exact verdict for every configuration in between.

Why is making a video classified as Media generation rather than a different task class?

A produced video is manufactured from a specification rather than retrieved or transformed from an existing source, which is exactly the property that defines the media generation class.

Would a longer or more complex video brief change the verdict shown here?

A different brief produces a different task hash, and a different task hash produces a different verdict — but this page publishes exactly one canonical brief per family, fixed in the Registry's own records, so the table above answers for that brief alone. A buyer's own brief is assessed separately, on order.