The Laugh Now Cry Later AI trend is an 11-second AI remake of the studio portrait from Drake and Lil Durk's 2020 music video, in which two people hold basketballs, one wags a finger at the camera and the other stays stone-faced. You can make it from two photos, or one photo of you both, with no prompt.
The page covers the basics. This guide goes one step deeper: where the AI version started, which bit of the song the sound uses, how it differs from the real video, how to choose photos for a duo, what it costs and how to post it safely.
Key takeaways
- The trend sound is an 11-second piece of Lil Durk's verse, not the chorus, and it has no explicit lyrics.
- The AI version began with the Summrs app's TikTok post of 23 September 2026 and is fading as of November 2026.
- Two adults, two photos, no hugging or touching, and mouths are not lip-synced.
- A Standard 720p 11-second video is 10 credits, and packs start at $6.99.
- Platforms may mute the song, so keep a licensed sound as a fallback.
In this guide
- Where did the AI version come from?
- Which part of the song does the trend sound use?
- How is it different from the real 2020 music video?
- Why the finger wag and the stone face?
- What happens second by second?
- How do I pick photos for a duo?
- Will we hug or touch in the video?
- Summrs app or a web generator: which should I use?
- What does a video cost?
- How do I post without getting the sound muted?
- Is it still worth making now that it is fading?
Where did the AI version come from?
The earliest AI post we found is the Summrs app's own TikTok post of 23 September 2026, made with its template and carrying a made-with badge. It had about 340,000 views. We could not confirm an individual creator behind it, so treat any "who started it" claim with caution.
Tutorial accounts were showing how to copy the look from mid October 2026, and CapCut template copies followed. Remakes with friends, couples and famous duos came after that. As of November 2026 new posts have slowed and the trend is fading, which matters for how you plan a post (see the last question).
Which part of the song does the trend sound use?
Not the chorus. The template uses an 11-second piece of Lil Durk's verse, roughly 1:27 to 1:38 of the track, and that stretch has no explicit lyrics. That is one reason the sound travels well: it is short, rhythmic and safe to put under a clip.
The song itself is Drake's "Laugh Now Cry Later" featuring Lil Durk, released on 14 August 2020. According to Wikipedia it peaked at number two on the US Billboard Hot 100 and number one on the Canadian Hot 100. Because the sound is a cut of the original recording, the finished video carries it, and the platform you post on decides whether it stays audible.
How is it different from the real 2020 music video?
The official video was directed by Dave Meyers and, per Wikipedia, shot at Nike's headquarters in Beaverton, Oregon. The AI trend borrows only one image from it: the dark studio portrait with two people holding basketballs. It is a recreation of a look, not a shot-for-shot copy of the video.
The videosi version is its own 11-second sequence. It opens on the finger wag, flashes a film-strip of full-length poses, lets the second person step forward, and ends on a night jet-ski shot across dark water. Your clothes come from your photos, and nobody from the original pair is kept in the result.
Why the finger wag and the stone face?
It is a two-part joke. One person delivers a smug, teasing gesture straight to the lens and the other refuses to react, so the clip reads as a status game between two people even with no dialogue. The contrast is what makes it funny when the pair are friends or partners who obviously know each other.
That is also why the two photo slots are labelled by role. The left person is the finger-wagger and the right person is the stone face. In the second half the roles shift a little: the right person steps forward and gestures at the camera while the left one stands behind with a ball, still stone-faced.
What happens second by second?
The clip runs 11 seconds in five beats:
- 0 to 2.5 s: close two-shot, the left person wags a finger while the right looks down at a basketball.
- 2.5 to 4 s: a contact-sheet transition flashes full-length poses.
- 4 to 5.5 s: wide shot of both standing side by side as the right person starts walking forward.
- 5.5 to 9.5 s: the right person gestures at the lens while the left stands behind.
- 9.5 to 11 s: cut to dark water at night and a jet ski past lit buildings.
Knowing the order helps you choose photos, because the wag and the jet ski each feature one person more than the other.
How do I pick photos for a duo?
Use two front-facing photos with head and shoulders in even light, one per person, or a single photo that shows both of you clearly. Avoid sunglasses, hats that hide the face, tight face crops and dark images. These are the usual reasons a result does not look like you.
Decide who plays which side before you upload. Photo 1 is the finger-wagger on the left; photo 2 is the stone face on the right. If you use one shared photo, put the finger-wagger on the left of the frame. A photo with more than two faces can confuse the clip, since it has exactly two adults.
Will we hug or touch in the video?
No. The two people never hug, kiss or touch in the videosi version, which keeps it usable for friends, siblings, coworkers and couples alike. Each person keeps their own face, hair, skin tone and build, and the two faces are tied to their own side so they do not blend.
Mouths move with the scene but are not lip-synced to the song, so do not expect it to look like you are rapping the verse. Likeness is also not guaranteed across all 11 seconds, especially in wide shots.
Summrs app or a web generator: which should I use?
Summrs started the trend, and its template is the origin of the look. A web generator such as videosi is a different route to the same format: you upload two photos in a browser, choose 9:16, 1:1 or 16:9, press Generate and download an 11-second MP4 without a watermark and with the trend sound already mixed in. videosi is independent and not affiliated with Summrs.
The practical differences are the inputs and the output. On the web you can supply two separate photos, choose a video quality tier, and pay per video through credits, instead of working inside one app's template. If you only want a quick in-app post and already use that app, its template is fine.
What does a video cost?
It is not free; there is no free generation. A Standard 720p 11-second video is 10 credits, and better models and higher resolutions cost more. Credit packs start at $6.99 for 10 credits, and a failed, timed-out or blocked render returns its credits automatically.
If you want to compare tiers, the effect page lets you pick the model before you generate. Standard is the sensible first try for a trend that is already past its peak.
How do I post without getting the sound muted?
The clip carries 11 seconds of the Drake and Lil Durk recording, and platforms may mute or limit reach for copyrighted audio. If that happens, upload the video without its sound and add the same track from the platform's own sound library, which is licensed for use there.
Also label the clip as AI-generated where the platform offers a toggle, and only upload photos of adults who agreed to it. Never use photos of minors. Many posts used famous duos, but you need the right to use any photo you upload.
Is it still worth making now that it is fading?
The peak was in October 2026, and by November new posts had slowed. A fading trend still works for personal posts, group chats and a birthday or couple clip, where the point is the joke between the two of you rather than reach.
If you want views, be early on the next trend instead. For this one, keep it light: one try with good photos beats several rushed ones, and each try uses credits. You can start from the Laugh Now Cry Later AI page whenever you are ready.
