Do I need to film anything or hire a creator?

No. The Avatar square opens a built-in catalogue of faces, plus anything you have uploaded to this project and the avatar pinned in your Reference Pack. Whatever you pick becomes the image-to-video seed the clip is rendered from. Uploads join the project corpus, so the same face is one Browse click away next run. Product is the one square with no stock catalogue: it takes your uploads and the pack pin only.

How long can a clip be and what size does it come out?

The Duration slider runs 5 to 15 seconds and opens at 10. Extended, the multi-scene mode, sits in the Model control as a disabled option marked coming soon, capped to a single 15 second block. Aspect is 9:16 at 1080x1920 or 16:9 at 1920x1080, with resolution locked at 1080p instead of exposed as a control. Generate makes one clip and stays blocked until that clip lands.

What do I have to write?

Nothing, if you would rather not. Script direction is a free text box whose placeholder shows the shape it wants: a Black Friday push for construction PMs, ROI number first, booking CTA last. Leave it blank and the render works from your brand voice, or press Fill with AI, which drafts one and saves it to the Script library below for reuse. Tone offers ten registers, Formal through Bold / edgy, with Neutral as the default.

What does it do with my brand assets?

If the project has a Reference Pack, its avatar, background and product sit first in each square and come preselected, which is what keeps a run of clips reading as the same person in the same room. With no pack pinned, a banner warns that clips drift across runs and links to Brand Identity, Reference Pack. Picking something else overrides the pack for that one clip; promote it into the pack to make it stick.

Why does the output look handheld instead of polished?

That is written into the render instructions rather than left to chance. The prompt refuses studio lighting, softboxes and ring lights, color-graded or cinematic LUT looks, shallow depth-of-field beauty-shot bokeh, voiceover-narrator delivery and stock-footage B-roll inserts. How rough it goes is yours: Camera spans twelve looks from Phone front cam (raw) to Body cam, CCTV night and Old Sony Ericsson, and Shake runs Locked (tripod), Subtle breath, Natural handheld, Vlogger walk-and-talk, Chaotic / running.

How long does a render take and where does the clip go?

The Model control quotes about 45 seconds for Standard, and the queued toast repeats that number. The moment you press Generate UGC clip a card appears in the generation history rail on the right, then turns into a playable video once the render lands. Failed runs keep their card and print the reason, so a hung provider is reported as a timeout you can retry instead of a card that quietly vanishes.

Fresh out of the oven

These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it.

Video tool

AI UGC video generator

Pinnora's AI UGC video generator renders one creator-style talking-head clip per run. Three slots sit across the top: Avatar, Background and Product. Avatar and Background each open a built-in catalogue; Product is fed only by your own uploads and the Reference Pack pin. The avatar you pick is the image-to-video seed the clip is built from, so nothing gets filmed. Below the squares a chip rail sets Model, Intent (Auto by default, or hook validation, objection buster, founder story, demo, social proof, comparison, POV trend, retargeting), Camera, Shake, Sound, Style, Tone and Cadence. A Duration slider covers 5 to 15 seconds, Aspect is 9:16 or 16:9 locked at 1080p, and On-screen captions burns TikTok-style text into the frame.

What you get

Produces: ugc talking head, video
Input: Starts from a written brief. No upload needed.
Price: Free to try. Sign in when you generate.
Category: Video

How it works

Four steps, start to finish

The tool is open above. This is the short version of what to do with it.

01 / 04
Fill the three squares
Avatar and Background open a built-in catalogue; Product takes only your uploads and the Reference Pack pin. Hover any square for its Browse and Upload buttons, and an upload joins the project corpus for later clips. A pack pin, or the first catalogue item, comes preselected.
02 / 04
Dial the shot
Camera covers twelve looks, Phone front cam (raw) through Dashcam, CCTV night and Old Sony Ericsson. Shake goes Locked (tripod) to Chaotic / running. Sound picks Ambient room, Voice-over only, Music bed or Silent (text-only). The Duration slider covers 5 to 15 seconds.
03 / 04
Set the read
Intent sets the angle and defaults to Auto: hook validation, objection buster, founder story, demo, social proof, comparison, POV trend or retargeting. Tone runs Formal through Bold / edgy, Neutral by default. Cadence is Calm and confident, Natural or Quick and direct. Leave Script direction blank, or press Fill with AI.
04 / 04
Render and check
Choose 9:16 or 16:9, leave On-screen captions on for burned-in TikTok-style text, then press Generate UGC clip. A card lands in the history rail on the right while it renders, then plays there. A failed run keeps its card and prints the reason.

FAQ

Common questions

Short answers to what people ask before they start.

Do I need to film anything or hire a creator?: No. The Avatar square opens a built-in catalogue of faces, plus anything you have uploaded to this project and the avatar pinned in your Reference Pack. Whatever you pick becomes the image-to-video seed the clip is rendered from. Uploads join the project corpus, so the same face is one Browse click away next run. Product is the one square with no stock catalogue: it takes your uploads and the pack pin only.
How long can a clip be and what size does it come out?: The Duration slider runs 5 to 15 seconds and opens at 10. Extended, the multi-scene mode, sits in the Model control as a disabled option marked coming soon, capped to a single 15 second block. Aspect is 9:16 at 1080x1920 or 16:9 at 1920x1080, with resolution locked at 1080p instead of exposed as a control. Generate makes one clip and stays blocked until that clip lands.
What do I have to write?: Nothing, if you would rather not. Script direction is a free text box whose placeholder shows the shape it wants: a Black Friday push for construction PMs, ROI number first, booking CTA last. Leave it blank and the render works from your brand voice, or press Fill with AI, which drafts one and saves it to the Script library below for reuse. Tone offers ten registers, Formal through Bold / edgy, with Neutral as the default.
What does it do with my brand assets?: If the project has a Reference Pack, its avatar, background and product sit first in each square and come preselected, which is what keeps a run of clips reading as the same person in the same room. With no pack pinned, a banner warns that clips drift across runs and links to Brand Identity, Reference Pack. Picking something else overrides the pack for that one clip; promote it into the pack to make it stick.
Why does the output look handheld instead of polished?: That is written into the render instructions rather than left to chance. The prompt refuses studio lighting, softboxes and ring lights, color-graded or cinematic LUT looks, shallow depth-of-field beauty-shot bokeh, voiceover-narrator delivery and stock-footage B-roll inserts. How rough it goes is yours: Camera spans twelve looks from Phone front cam (raw) to Body cam, CCTV night and Old Sony Ericsson, and Shake runs Locked (tripod), Subtle breath, Natural handheld, Vlogger walk-and-talk, Chaotic / running.
How long does a render take and where does the clip go?: The Model control quotes about 45 seconds for Standard, and the queued toast repeats that number. The moment you press Generate UGC clip a card appears in the generation history rail on the right, then turns into a playable video once the render lands. Failed runs keep their card and print the reason, so a hung provider is reported as a timeout you can retry instead of a card that quietly vanishes.

Academy course

Brand setup

Your product, your people and your palette show up in the work, instead of a stock stand-in.

3 lessons · 15 minStart the course →

More tools

Related tools

The rest of the family. Every one is free to try and runs on the same engine.

All 24tools →

Fresh out of the oven

These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it.

Fresh out of the oven

These tools are new, so the odd thing may still wobble. If something looks wrong, it probably is. Tell us and we will sort it.

Video tool

AI UGC video generator

What you get

Produces: ugc talking head, video
Input: Starts from a written brief. No upload needed.
Price: Free to try. Sign in when you generate.
Category: Video

How it works

Four steps, start to finish

The tool is open above. This is the short version of what to do with it.

01 / 04
Fill the three squares
Avatar and Background open a built-in catalogue; Product takes only your uploads and the Reference Pack pin. Hover any square for its Browse and Upload buttons, and an upload joins the project corpus for later clips. A pack pin, or the first catalogue item, comes preselected.
02 / 04
Dial the shot
Camera covers twelve looks, Phone front cam (raw) through Dashcam, CCTV night and Old Sony Ericsson. Shake goes Locked (tripod) to Chaotic / running. Sound picks Ambient room, Voice-over only, Music bed or Silent (text-only). The Duration slider covers 5 to 15 seconds.
03 / 04
Set the read
Intent sets the angle and defaults to Auto: hook validation, objection buster, founder story, demo, social proof, comparison, POV trend or retargeting. Tone runs Formal through Bold / edgy, Neutral by default. Cadence is Calm and confident, Natural or Quick and direct. Leave Script direction blank, or press Fill with AI.
04 / 04
Render and check
Choose 9:16 or 16:9, leave On-screen captions on for burned-in TikTok-style text, then press Generate UGC clip. A card lands in the history rail on the right while it renders, then plays there. A failed run keeps its card and prints the reason.

FAQ

Common questions

Short answers to what people ask before they start.

Do I need to film anything or hire a creator?: No. The Avatar square opens a built-in catalogue of faces, plus anything you have uploaded to this project and the avatar pinned in your Reference Pack. Whatever you pick becomes the image-to-video seed the clip is rendered from. Uploads join the project corpus, so the same face is one Browse click away next run. Product is the one square with no stock catalogue: it takes your uploads and the pack pin only.
How long can a clip be and what size does it come out?: The Duration slider runs 5 to 15 seconds and opens at 10. Extended, the multi-scene mode, sits in the Model control as a disabled option marked coming soon, capped to a single 15 second block. Aspect is 9:16 at 1080x1920 or 16:9 at 1920x1080, with resolution locked at 1080p instead of exposed as a control. Generate makes one clip and stays blocked until that clip lands.
What do I have to write?: Nothing, if you would rather not. Script direction is a free text box whose placeholder shows the shape it wants: a Black Friday push for construction PMs, ROI number first, booking CTA last. Leave it blank and the render works from your brand voice, or press Fill with AI, which drafts one and saves it to the Script library below for reuse. Tone offers ten registers, Formal through Bold / edgy, with Neutral as the default.
What does it do with my brand assets?: If the project has a Reference Pack, its avatar, background and product sit first in each square and come preselected, which is what keeps a run of clips reading as the same person in the same room. With no pack pinned, a banner warns that clips drift across runs and links to Brand Identity, Reference Pack. Picking something else overrides the pack for that one clip; promote it into the pack to make it stick.
Why does the output look handheld instead of polished?: That is written into the render instructions rather than left to chance. The prompt refuses studio lighting, softboxes and ring lights, color-graded or cinematic LUT looks, shallow depth-of-field beauty-shot bokeh, voiceover-narrator delivery and stock-footage B-roll inserts. How rough it goes is yours: Camera spans twelve looks from Phone front cam (raw) to Body cam, CCTV night and Old Sony Ericsson, and Shake runs Locked (tripod), Subtle breath, Natural handheld, Vlogger walk-and-talk, Chaotic / running.
How long does a render take and where does the clip go?: The Model control quotes about 45 seconds for Standard, and the queued toast repeats that number. The moment you press Generate UGC clip a card appears in the generation history rail on the right, then turns into a playable video once the render lands. Failed runs keep their card and print the reason, so a hung provider is reported as a timeout you can retry instead of a card that quietly vanishes.

Academy course

Brand setup

Your product, your people and your palette show up in the work, instead of a stock stand-in.

3 lessons · 15 minStart the course →

More tools

Related tools

The rest of the family. Every one is free to try and runs on the same engine.

All 24tools →

AI UGC video generator

Fill the three squares

Dial the shot

Set the read

Render and check

Brand setup

AI UGC video generator

Fill the three squares

Dial the shot

Set the read

Render and check

Brand setup