Invisible moped AI for two: one rides, one gives chase

Updated 2026-10-19

Three two-person chases and one solo chase from other generators, credited under each video. They follow the same overpass-chase motion as this generator, but this site did not make them and your photos will look different. Their lengths vary; this tool makes 15 seconds.

Photo 1 rides and photo 2 chases. This is how pairs usually split the roles, and what to know for each.

Who is on screen, second by second

1. 1. Pick one photo per person

   Use a photo of each person with nobody else in it, face toward the camera, no sunglasses. The two photos can come from different days and places; the generator puts both people into one scene. Upload the rider first and the chaser second, and press Swap if you change your mind.

2. 2. Choose the place, frame and model

   Keep the overpass, or pick neon city, the supermarket or an empty photo of your own place. 9:16 is set for Stories, Reels and TikTok; switch to 16:9 for YouTube. The credit quote updates as you change the model or quality.

3. 3. Generate, check both faces, download

   Sign in and generate; you can leave the page while it runs. Check the chaser’s close-up at about 5 seconds and the rider through the flip, then download the MP4. For the reverse version, press Swap and generate again as a new run.

These problems are specific to two-person chases. The fixes are changes to your photos and roles; no setting guarantees a perfect result.

A video that finished but looks wrong is a new paid run when you generate again. Only failed tasks return their credits.

If you build the chase yourself in a general video model, this prompt keeps two people apart: each role is tied to one reference image, and the timeline gives the chaser a close-up of their own. Attach the rider’s photo as image 1 and the chaser’s as image 2.

Vertical 9:16, 15 seconds, photorealistic. Two people from two reference photos: Image 1 is the RIDER, Image 2 is the CHASER. Keep each person's own face, hair, skin tone, glasses and build; never give one person the other's face.
0–5 s: the RIDER crouches in mid-air on an invisible moped on a concrete overpass at dusk, knees bent, fists forward on handlebars that are not there, grinning. Far behind, the CHASER appears on a white wheelless hover-bike.
5–7 s: close-up of the CHASER alone, in a grey police uniform and peaked cap, frowning into the lens, hands on the hover-bike's handlebars.
7–9 s: the RIDER looks back over one shoulder at the CHASER, now close behind.
9–11 s: the RIDER does one clean backflip and lands back in the crouch, both hands up, shouting with joy.
11–15 s: low wide shot, the RIDER shoots away down the road, the CHASER following, both small in frame.
No vehicle, seat or wheels under the RIDER at any moment. One RIDER and one CHASER only, no extra people, no text or watermark.

Written for this page and not the generator’s own template prompt; the generator above needs no prompt. Results depend on the tool you paste it into.

Two photos cost the same as one: 162 credits for the default (Seedance 2.5, 480p), worth $9.43 from a Single Reveal pack or $4.49 at the Creator Pack rate. Your first checkout is still $9.90 for 170 credits, not the per-video figure.

One-time packs; no renewal. Credits expire 365 days after purchase.

One chase for each pair in your group

Making a video for every couple or pair of friends? Each row starts from a zero balance and buys the cheapest pack mix. Failed tasks are refunded and not counted here; redoing a finished video is.

Ask both people. The service accepts photos of yourself or of adults who agreed to this use. That includes a coworker’s photo from a company website and a friend’s photo for a surprise: ask for the photo, and keep the video as the surprise.

Adults only. Photos of anyone under 18 cannot be uploaded, so a younger brother, sister or child cannot ride or chase. The video provider or our own checks may decline a photo that looks like a minor; declined tasks return their credits.

Storage and posting. Photos and videos are kept so you can download them from My videos and are deleted when you email support@unforgettablecarai.app; they are not used to train AI models. On TikTok, turn on the AI-generated content label when you post a realistic AI video.

The name points back to Crazy Frog’s 2005 “Axel F” video, a UK number one in which the animated frog rides an invisible motorbike and outruns a futuristic pursuer. In October 2026 AI remakes put real people in that pose, and a clip posted by @wdmowski on 12 October 2026 (a man crouched over an overpass, chased by a hover-bike cop) set the shots most remakes copy. It is a traceable early example, not proof of who started the trend.

This generator follows that overpass chase: the same shots, cuts and timing, with your two photos as the cast. The examples above come from other generators and are credited to them.

• Crazy Frog – Axel F (official video, YouTube)
• The overpass chase by @wdmowski (TikTok, 12 October 2026)