# Pick One Video Generator

> The Let Go trend, generated. Type a topic; AutoFeed writes the hook, renders three to ten photorealistic options and cuts them to the beat of the Let Go track. Viewers reply with a number.

- Canonical page: https://www.autofeed.ai/generators/pick-one-video-generator
- Publisher: AutoFeed.ai (https://www.autofeed.ai)
- Site overview for language models: https://www.autofeed.ai/llms.txt

**Best for:** Fast, silent, comment-bait uploads with almost no writing.

## What a Pick One video is

A Pick One video — also called the Let Go or VibePlug format — opens on a hook clip carrying two lines of text, stacked as one centred block: an absurd scenario, and under it the choice question, which stays on screen over every image that follows. Then it runs through numbered images, one every few seconds: rooms, pools, sinks, cabins. There is no narration and no script to follow. The whole format is a comment machine, because the ask is a single number, which is the lowest-effort interaction available on any platform.

AutoFeed writes the hook and the image prompts from one topic line, generates the images, picks a hook clip at random from a library of 50 and cuts the timing to whichever track you chose. There is no art-style control here and that is deliberate: the image prompt pins one aesthetic — editorial architecture and interiors photography, golden-hour, blue-hour or warm tungsten light, f/1.4 to f/2.8 depth of field, a little film grain, switching to product photography when the options are objects rather than spaces. The contrast is the point: the text describes something chaotic or gross, and the images are immaculate.

## Key specifications

| Spec | Value |
| --- | --- |
| Options per video | 3-10 images, five by default |
| Hook clips | 50, one picked at random |
| Voiceover | None — silent by design |
| Credits | 1,000 + 100 per image, +400 if animated |
| Length at 5 images | ~19s on the default track |

## Example videos

- A Pick One video generated by AutoFeed — hook clip, scenario and question, then numbered options. (https://www.autofeed.ai/landing/videos/pick-1.mp4)

## How the Pick One generator works

1. **Type a topic** — One line is enough. The box is pre-filled from a pool of nine — "Beautiful showers from around the world", "Cabins to disappear into for a whole winter", "Airport lounges you would happily miss your flight in" — and the pool advances one step per UTC day, so the pre-fill you see today is the same one everybody else sees today and a different one tomorrow. The rotation is keyed to the date rather than randomised so the server and the browser always agree on what is in the box. Overwrite it with anything you want; the point of the rotation is only that nobody is nudged toward the same subject every single day.
2. **Pick how many options** — Three to ten images. That number is the video: on the default slowed track the hook runs 3.7 seconds and each image holds 2.9, so five options land at about 19 seconds and ten at about 33.
3. **Generate the hook and the prompts** — The model returns a title, a description with hashtags, a scenario line ending in an emoji, a question line, and one image prompt per option. The two lines render as one centred block in the middle of the opening frame, not pinned to the edges. Both are capped by character count rather than word count, because that is what decides how many rendered lines they take: 70 characters for the scenario, 45 for the question. Both are plain text boxes you can rewrite.
4. **Decide how it moves** — Zoom Effect alternates a slow push in and out across the stills. Animated Images instead converts every still into a five-second 720p clip; that costs an extra 400 credits per image — so 1,200 more at three images and 4,000 more at ten — takes longer, and switches the zoom off because the clip already moves.
5. **Match the music** — Pick a track from the 87-track sound library. Let Go (Slowed) is the default, and it and Let Go (Guitar) are the two cuts the edit is beat-matched to: 3.7 seconds of hook and 2.9 per image on the slowed version, 2.52 and 2.4 on the guitar one. Any other track falls back to a flat 3-second hook and 3 seconds per image.
6. **Render, then let it run** — Export at 9:16, 16:9 or 1:1. Download it, or run the format as a series that publishes to TikTok and YouTube Shorts automatically.

## Everything you can configure

### Topic

The one line everything else is written from.

- Any theme — rooms, places, objects, food
- Pre-filled from a pool of nine, advancing one step per UTC day
- Same pre-fill for everyone on a given day; a different one tomorrow

### Number of images

How many options the viewer chooses between.

- 3 to 10 per video, five by default
- The editor shows the resulting length as you change it
- Roughly 19 seconds at five images on the default track

### Hook scenario and question

The two lines of text on the opening clip.

- Both render as one centred block, not pinned to the edges
- Scenario — at most 70 characters, ends in an emoji
- Question — at most 45 characters, and it stays on screen over every image
- Both are editable text boxes

### Movement

Two mutually exclusive ways to stop the stills feeling static.

- Zoom Effect — alternating push in and out, on by default
- Animated Images — each still becomes a five-second 720p clip
- Animated adds 400 credits per image, on top of the 100 the image costs
- Turning on Animated Images disables the zoom

### Music

The track, which also sets the edit timing.

- 87-track library, one track per video
- Let Go (Slowed), the default — 3.7s hook, 2.9s per image
- Let Go (Guitar) — 2.52s hook, 2.4s per image
- Any other track — 3s hook, 3s per image

### Aspect ratio

The shape of the finished file.

- 9:16 vertical
- 16:9 horizontal
- 1:1 square

## 24 Pick One Hooks, With the Topic That Feeds Them

The trend runs on contrast: the text describes something chaotic, gross or nostalgic, and the images are immaculate. Each entry below is a topic to generate from plus the scenario and question that should sit on the hook — rewrite either line once the images come back. Note how few of them open with "POV:". The generator is held to at most one in four for exactly that reason: a feed of identical openers stops reading as a video and starts reading as a template.

*Where to paste:* Paste the topic into the Topic box on the Content tab, then press Edit on the post and paste the two lines into the Scenario and Question boxes.

### Hygiene and reset

The highest-performing category on the trend. The setup is always deprivation; the payoff is always spotless.

- Marble bathrooms with rain showers — Scenario: POV: you've been camping in the rain for nine days 🌧️ · Question: Which shower are you taking first?
- Hotel bathtubs with a city view — Scenario: Your flight got cancelled three times today ✈️ · Question: Which bath are you sinking into?
- Hotel beds with mountains of pillows — Scenario: You've slept on a friend's floor all month 😴 · Question: Which bed are you claiming?
- Infinity pools at sunset — Scenario: A week with no air conditioning and no end in sight 🥵 · Question: Which pool are you jumping into?
- Steamy Japanese onsen at night — Scenario: You walked home in the snow without a coat ❄️ · Question: Which bath are you getting into?
- Laundry rooms that look like showrooms — Scenario: Same hoodie since Tuesday. It's Saturday 😭 · Question: Where are you washing everything?

### Absurd and chaotic

The more hyperbolic the setup, the more comments. The images should stay completely serious.

- Cheap-looking mansions that are secretly enormous — Scenario: POV: your dog signed a lease while you were asleep 🐕 · Question: Which house are you living in now?
- Treehouses built into impossible cliffs — Scenario: The ground floor is lava now. Permanently 🔥 · Question: Which treehouse are you climbing into?
- Underground bunkers that look like hotels — Scenario: You agreed to house-sit. The address was underground 🚪 · Question: Which room are you taking?
- Abandoned malls turned into apartments — Scenario: You're the last person on earth and it's all free 🌍 · Question: Which mall are you moving into?
- Boats you could actually live on — Scenario: Your street floated away overnight 🌊 · Question: Which boat are you swimming to?
- Castles that are somehow for sale — Scenario: A stranger left you a castle in their will 🏰 · Question: Which one are you keeping?

### Nostalgia and time travel

Aim at a specific year, not a vague decade. Precision is what makes the comments say "this is exactly it".

- 2000s bedrooms with CRT TVs and fairy lights — Scenario: POV: it's a Friday night in 2007 and nobody can reach you 📼 · Question: Which room are you falling asleep in?
- Empty arcades at midnight — Scenario: You're twelve again and the tokens never run out 🕹️ · Question: Which arcade are you never leaving?
- Video rental stores that never closed — Scenario: Streaming was never invented 📼 · Question: Which aisle are you starting in?
- Summer camp cabins by a lake — Scenario: School just ended and it's the first week of July ☀️ · Question: Which cabin is yours?
- Grandparents’ kitchens, one from every decade — Scenario: It's Sunday and dinner is already on 🍲 · Question: Which kitchen are you sitting in?
- Old libraries with green desk lamps — Scenario: Your phone died and nobody knows where you are 📵 · Question: Which desk are you working at?

### Hyper-specific POV

The narrowest setups travel furthest. "A long flight" is nothing; "a fourteen-hour layover" is a comment section.

- Airport lounges with floor-to-ceiling windows — Scenario: POV: your layover is fourteen hours long ✈️ · Question: Which lounge are you camping in?
- Cafés with rain running down the window — Scenario: It's raining and you have absolutely nothing due 🌧️ · Question: Which window are you sitting by?
- Fireplaces in mountain cabins — Scenario: Four hours of walking in the freezing cold 🥶 · Question: Which fire are you sitting at?
- Home cinemas with terrible taste and great seats — Scenario: It's 2 AM and you are not remotely tired 🎬 · Question: Which screen are you watching?
- Balconies over a city that never sleeps — Scenario: You can't sleep and the flat is too quiet 🌃 · Question: Which balcony are you standing on?
- Desert road diners at golden hour — Scenario: Eleven hours of driving and it's finally dark 🚗 · Question: Which diner are you pulling into?

## Frequently asked questions

### Why is there no voiceover?

Because the trend does not have one. The format is built around a music bed and two lines of on-screen text, and adding narration is the quickest way to make it look like an imitation. AutoFeed skips voice generation entirely for this format — there is no voice control in the editor at all.

### How long does a Pick One video end up being?

The image count and the track decide it together. On the default slowed cut the hook runs 3.7 seconds and each image holds 2.9, so five options is roughly 19 seconds and ten is roughly 33. The editor shows the number as you change the count.

### What is the hook clip?

A short piece of footage that the scenario and question lines sit on top of. AutoFeed keeps a library of 50 and picks one at random on every render, so consecutive uploads in a series do not all open identically.

### Should I turn on Animated Images?

Only when the stills are already strong. It converts each image into a five-second 720p clip at 400 credits apiece — a five-image video goes from 1,500 credits to 3,500 — adds render time, and disables the zoom effect. For most Pick One videos the alternating zoom on sharp stills reads better than motion for its own sake.

### Can I rewrite the hook text?

Yes — the scenario line and the question line are separate boxes. Keep the scenario inside 70 characters and the question inside 45, which is what the generator itself is held to: at the size they are burned in, roughly 55 characters of scenario fills two rendered lines and 70 fills three, and the question is redrawn over every image afterwards, so it has to read on its own.

### How much does one cost?

A thousand credits plus a hundred per image, so the range runs from 1,300 at three images to 2,000 at ten, and the five-image default is 1,500. Animated Images adds a further 400 per image: 2,500 at three, 3,500 at five, 6,000 at ten.

### Can I change the art style?

No, and that is on purpose — Pick One is the one format with no style picker. The image prompt pins a single aesthetic: editorial architecture and interiors photography, golden-hour, blue-hour or warm tungsten light, f/1.4 to f/2.8 depth of field and a little film grain, switching to product photography when the options are objects rather than rooms. It is what the trend actually looks like, and a cartoon or anime version of it reads as a different format entirely.

### Why do the generated images never contain numbers?

The prompt explicitly forbids text, numbers, labels and watermarks inside the image, because the numbers are drawn over the top at render time. Baked-in text would fight the overlay and, in practice, comes out garbled.

## Related

- [AI Video Generator](https://www.autofeed.ai/features/ai-video-generator)
- [All generators](https://www.autofeed.ai/generators)
- [Pricing](https://www.autofeed.ai/pricing)