Checklist the per-video bar — engine/sim
86.6/100
quantitative quality · weights learn from your reviews (
engine.sim.memory review social-W29-Thu-3 good|bad)⚠ 3 flag(s) — not yet ship-ready: copy_genericdup_sequence_across_assetsdup_set_across_assets · see docs/strategy/VIDEO-CHECKLIST.md
Composition comp · template family · expected output
9:16 Reelpending1:1 Squarepending16:9 Widepending9:16 4Kpending1:1 4Kpending16:9 4KpendingGIF (SMS)pending
render pending — silent master not yet on disk
expected output: 0/7 rendered — same matrix the /media preview surfaces for this asset.
Composition layer × scene 9 scenes · 71.0s · comp_id + rendered still + tier + the script
| # | Layer (comp_id · still · tier) | Beat | Timecode | Motion | Logo | Audio | VO / on-screen / caption |
|---|---|---|---|---|---|---|---|
| 1 |  matches intent shared field signature-3d | open | 0–7.3s | spatial-parallax | icon·white-knockout | ♪ bed_in | Not long ago, doing anything with AI meant renting a giant server farm you could never afford to keep. on-screen: AI used to mean a giant server |
expected on screen: black ground · box hero in the shared Signal Field · Faber leads · node-graph · spatial-parallax · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape boxground blacktreatment white-knockoutmotion spatial-parallaxpower summoninstrument summon→AI used to mean a giant server farm | |||||||
| 2 |  matches intent shared field signature-3d | hook | 7.3–15.7s | kinetic-build | icon·ember-fill | ♪ node_lock | The wall in front of small teams was never talent. It was the bill that arrived at the end of the month. on-screen: The wall was cost, not ability |
expected on screen: black ground · box hero in the shared Signal Field · Faber leads · mark · kinetic-build · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual markshape boxground blacktreatment ember-fillmotion kinetic-buildpower laser-lockinstrument laser-trace→The wall was cost, not ability | |||||||
| 3 |  matches intent NumberedList template | teach | 15.7–23.799999999999997s | kinetic-build | icon·wireframe | ♪ node_lock | Now you send a model one request over the web and pay only for the words it reads and writes back. on-screen: Call a model with one request |
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id NumberedListvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Call a model with one requestcurate items, nodes | |||||||
| 4 |  matches intent ComparisonTable template | proof | 23.8–32.6s | receipts-count | icon·white-knockout | ♪ node_lock | You pay by the word instead of by the server, so a tiny team can run the same brain as a giant one. on-screen: Pay per word, not per server |
expected on screen: black ground · a ComparisonTable panel over a dimmed Signal Field · Faber leads · receipts · receipts-count · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id ComparisonTablevisual receiptsshape boxground blacktreatment white-knockoutmotion receipts-countpower receiptsinstrument spotlight→Pay per word, not per servercurate colA, colB, rows, statsLabels | |||||||
| 5 |  matches intent DiffCard template | diff | 32.6–41.0s | crossfade-8f | icon·liquid-chrome | ♪ node_lock | The Big T-M does not buy machines that sit idle. We pay for the answer and let someone else own the metal. on-screen: We pay for results, not for hardware |
expected on screen: black ground · a DiffCard panel over a dimmed Signal Field · Faber leads · code · crossfade-8f · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id DiffCardvisual codeshape boxground blacktreatment liquid-chromemotion crossfade-8fpower morphinstrument morph→We pay for results, not for hardwarecurate codeLines, lines | |||||||
| 6 |  matches intent ChecklistCard template | teach | 41.0–48.3s | kinetic-build | icon·ember-fill | ♪ node_lock | Save the answers you ask for again, so you never pay twice for the same question you already solved. on-screen: Cache the answers you reuse |
expected on screen: black ground · a ChecklistCard panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id ChecklistCardvisual node-graphshape boxground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Cache the answers you reusecurate items, nodes | |||||||
| 7 |  matches intent KpiGrid template | proof | 48.3–55.3s | receipts-count | icon·wireframe | ♪ node_lock | Caching turns a repeated question into a free lookup, and the monthly bill stops climbing for no reason. on-screen: The same question, paid for once |
expected on screen: black ground · a KpiGrid panel over a dimmed Signal Field · Faber leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id KpiGridvisual receiptsshape boxground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→The same question, paid for oncecurate kpis, statsLabels | |||||||
| 8 |  matches intent AnnotatedDiagram template | step | 55.3–62.599999999999994s | kinetic-build | icon·white-knockout | ♪ node_lock | Try it: make one small model request, save the result, and reuse it the next time the question repeats. on-screen: Do this |
expected on screen: black ground · a AnnotatedDiagram panel over a dimmed Signal Field · Faber leads · pipeline · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id AnnotatedDiagramvisual pipelineshape boxground blacktreatment white-knockoutmotion kinetic-buildpower throwinstrument laser-trace→Do this: one API call, then cache itcurate callouts, stages, steps | |||||||
| 9 |  matches intent shared field signature-3d | resolve | 62.6–71.0s | coalescence | icon·white-knockout | ♪ bed_out | The old AI wall fell quietly. The server farm did not get cheaper to own. It just stopped needing to be yours. on-screen: The farm came to you |
expected on screen: black ground · box hero in the shared Signal Field · Faber leads · coalescence · coalescence · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape boxground blacktreatment white-knockoutmotion coalescencepower coalescenceinstrument coalescence→The farm came to you | |||||||
Format stack 3 aspects · same scenes[], re-cropped
9:16
1080×1920
Stories · TikTok · YouTube Shorts · Reels
1:1
1080×1080
LinkedIn · Facebook · Instagram
16:9
1920×1080
X/Twitter · YouTube · LinkedIn video
Channels 9 destinations
LinkedInX/TwitterYouTubeInstagramFacebookThreadsTikTokPinterestBluesky
Social captions supplemental published copy · per channel (comp_id level)
tiktokAI used to mean renting a server farm you couldn't afford (AI-assisted breakdown). Now: send a model ONE request over the web, pay only per word it reads and writes. The move? Cache answers you reuse so you never pay twice for the same question. Pay for results, not hardware. #ai #api #tech
instagramThe AI wall was cost, not talent.
Call a model with one request.
Pay per word, not per server.
Cache the answers you reuse.
Never pay twice for the same question.
#ai #automation #api #tech #smallbusiness
linkedinAI and automation, the cost story nobody tells. Not long ago, doing anything with AI meant renting a server farm a small team could never afford; the wall was the bill, not the talent. Now you send a model one request over the web and pay only for the words it reads and writes back, so a tiny team runs the same brain as a giant one. The operator move: cache the answers you reuse, so a repeated question becomes a free lookup and the bill stops climbing for no reason. Pay for results, not hardware.
xAI used to mean renting a server farm you couldn't afford. Now: send a model ONE request, pay per word. The move: cache answers you reuse so you never pay twice for the same question. Pay for results, not hardware.
facebookNot long ago, doing anything with AI meant renting a giant server farm you could never afford. The wall was the bill, not the talent. Now you send a model one request over the web and pay only for the words it reads and writes back. The smart move: cache the answers you reuse so you never pay twice for the same question. Pay for results, not hardware. More in the video.
threadsAI used to mean renting a server farm you couldn't afford. Now: send a model ONE request, pay per word. The move: cache answers you reuse so you never pay twice for the same question. Pay for results, not hardware.
pinterestAI on a budget: pay-per-token API, cache AI responses, small team AI, no server farm, cost control for AI, model API basics, save money on AI. How a tiny team can run the same brain as a giant one and stop paying twice.
blueskyAI used to mean renting a server farm you couldn't afford. Now: send a model ONE request, pay per word. The move: cache answers you reuse so you never pay twice for the same question.
youtubeAI & Automation: Pay Per Word, Cache What You Reuse
Not long ago, doing anything with AI meant renting a giant server farm a small team could never afford; the wall was the bill, not the talent. Now you send a model one request over the web and pay only for the words it reads and writes back, so a tiny team runs the same brain as a giant one. The operator move: cache the answers you reuse, so a repeated question becomes a free lookup and the monthly bill stops climbing. Pay for results, not hardware. AI-assisted, human-judged.