All guides

How to Make a Goodbye 2026 Welcome 2027 AI Video

November 6, 2026 · AI Videographer Team

A goodbye 2026 welcome 2027 AI video is a short New Year clip made from one photo. In the version on this site, two copies of you meet on a rainy night road under two overhead signs, 2026 on one side and 2027 on the other. You share one last hug, your 2027 self picks up a suitcase, gets into a waiting car and drives away while your 2026 self stays behind and watches. It is an easy post for December 31 or January 1.

There are two ways to make one. The quick way is a template that builds the scene around your face and animates it. The manual way is to create the still picture yourself with an image model and then animate it with a video model. This guide covers both, with the prompts that work, and what to do when the numbers or the faces come out wrong.

Open the Goodbye 2026 Welcome 2027 template

What the finished video looks like

The template makes a vertical 9:16 video at 720p, 10 seconds long, with rain and car sounds and no music. The demonstration on the template page was made from a photo of a fictional AI-generated adult. Your own video follows the same scene and action, but AI output varies from take to take.

The quick way: 5 steps

  1. Pick your photo. One adult, facing the camera, in good light. Show your outfit and as much of your body as you can, because the farewell scene is a medium full shot.
  2. Upload it to the template. Go to the Goodbye 2026 Welcome 2027 page and upload the photo. Use yourself or an adult who agreed; photos of celebrities and of people who did not agree are not allowed, and uploads are checked automatically.
  3. Check the cost. The template uses the 10-second 720p preset. The credit cost is shown before you start. There is no free generation preview.
  4. Sign in and create. Start the video and keep the page open while it renders. The tool first builds the rainy road picture with two copies of you, then animates the hug and the departure.
  5. Check the numbers and download. Watch the video and look at the 2026 and 2027 signs before you post. If they are right, download the MP4. If not, see the fixes below.

The manual way: build it in any AI tool

If you prefer your own tools, split the job in two. Asking a video model to invent a road, two signs, a car and two copies of the same face at once is where most attempts fall apart. Making the still picture first lets you check that the signs read 2026 and 2027 before you spend anything on the video.

Step one, the scene. Give your photo to an image editing model and describe the place, the two copies and the signs:

Use the single adult from the input photo to create exactly two identical copies of the same adult, preserving face, hair and clothing. They gently hug on a quiet rain-slick night road. Overhead signs read exactly 2026 on the left and 2027 on the right. A dark sedan and a small suitcase wait on the right. Cinematic natural rain, vertical 9:16, no other people, no logos, no extra text.

Step two, the motion. Give that picture to an image to video model and describe the action second by second:

From 0 to 2 seconds finish the hug. From 2 to 3 seconds the person on the right lets go and takes the suitcase. From 3 to 7 seconds that person walks to the car and gets inside. From 7 to 10 seconds the car drives away with red tail lights reflecting on the wet road while the person on the left stays and watches. Camera gently pulls back. Natural rain and engine sounds, no music, no dialogue.

A few rules make a big difference:

  • Write each year as one number and say it is the only text in the image. Extra words on signs are where models make spelling mistakes.
  • Say which side each copy stands on, so the 2026 you stays under the 2026 sign.
  • Keep the camera slow. Fast moves make the digits wobble and the faces drift.
  • Ask for exactly two people. Models like to add a stranger in the background.

How the two of you are made

The template uses one photo for two people. The image model draws the same face twice, and the motion prompt gives each copy a different job: one leaves, one stays. A neutral, front facing photo works best, because the model then has one clear face to copy and fewer details to invent.

When to post it

The farewell scene fits December 30 or 31, with a caption that thanks or roasts the old year. Make your video a few days early, so you are not waiting for a render at 11:55 pm. When the app asks, label the post as AI generated; TikTok, Instagram and YouTube all offer a label for AI made content.

When a take goes wrong

  • The signs read 2072 or a digit is bent. AI models still miss a digit sometimes. Run it again, or cover the sign with a 2027 sticker or caption in your editing app.
  • The face does not look like the photo. Use a sharper photo with the face turned to the camera, no sunglasses and no cap shading the eyes. A second run often looks closer.
  • The two copies look like different people. Use a front facing photo with a neutral expression and a plain background. Strong smiles or side angles make the second copy drift.
  • A third person appears. Run it again; the template asks for exactly two copies of one adult, but extra people still happen now and then.
  • The photo was rejected. Uploads are checked for nudity, violence, public figures and minors. Use a photo of yourself or of an adult who agreed. If a job fails or cannot pass review, its credits are restored.

If you are looking for this video by description

Many people search for this trend without knowing its name. All of these mean the same kind of video:

  • the goodbye 2026 hello 2027 AI video, or bye 2026 hi 2027
  • the AI video where you hug your old self under 2026 and 2027 road signs
  • the AI video where your 2027 self drives away in the rain
  • the New Year 2027 AI trend from one photo

FAQ

Can I make a goodbye 2026 welcome 2027 AI video for free? No. There is no free generation preview. The credit cost of the 10-second 720p preset is shown before you start; see pricing for credit packs, which are one-time purchases, not subscriptions.

How many photos do I need? One photo of one consenting adult. The AI creates both copies of that same person.

Does the video have sound? Yes. It uses rain and car ambience; there is no song and no dialogue.

Can I change the year on the signs? Not in the template, where the scene is built for 2026 and 2027. In the manual way above you can write any number into your own scene prompt.

Can I write my own prompt? The template has a built-in prompt, so you only upload a photo. For full control, use the manual way above in any image and video tool.

Try it with your photo

Sign in and check the credit cost before creating.

Create a video