The love name letter is a 15-second TikTok-style clip where the first letter of someone you love glows red beside a wall of "I love you" lines. You can make one from a single photo and one letter tap on the love name letter page, and the finished video comes with the trend's original sound. This guide answers the practical questions that come after the basics: which letter to pick, how to caption it, why AI struggles with text and what to expect from the result.
One honest note up front: this effect has not been test-rendered yet, so we cannot promise that any letter comes out exact. The answers below say what we know and where the limits are.
Key takeaways
- Pick the first letter of the name people actually use, or two letters for a couple with a small heart between them.
- The effect edits an existing letter clip instead of drawing a letter from nothing, which helps but does not guarantee an exact letter.
- Expect red light on your skin and clothes; it is the look of the trend.
- Get permission from anyone whose photo you use, and caption with a hint rather than a full name.
In this guide
- How do I choose the letter for a love name letter?
- Should I use a single initial or a couple initial?
- What should I write in the caption?
- Why does AI struggle with letters and text in video?
- How does this effect avoid scrambled letters?
- Has the effect been test-rendered, and will my letter be exact?
- Why does my skin and clothing turn red?
- Do I need permission to use someone else's photo?
- How is this different from making a letter edit in CapCut?
- Does the video come with sound?
- What does it cost?
How do I choose the letter for a love name letter?
Use the first letter of the name you would say out loud, not a nickname you only use in private. The trend works because viewers guess, so the letter should be one the person would recognise. If a person has two common names, pick the one friends use. The effect on videosi offers the 26 capital letters A to Z as a single tap, so there is nothing to type and no spelling to get wrong.
Should I use a single initial or a couple initial?
A single initial keeps the guessing game alive: one glowing letter, one mystery. A couple version shows two letters joined by a small red heart, which suits an anniversary or a public relationship post. On videosi the second letter is optional and defaults to none, so you can render the single version first and add a second letter for a different post. Both versions use the same 15-second timing.
What should I write in the caption?
Keep it short and let the letter do the work. The page suggests a caption such as "my love name letter", or tagging the person the letter is for and letting people guess. A line like "guess who" works for a single initial, and for a couple version you can simply ask which two names the letters belong to. Do not type the full name in the caption if the point is the hint.
Why does AI struggle with letters and text in video?
Image and video models learn shapes, not spelling, so strokes can bend, mirror or turn into look-alike symbols. Researchers describe the same issue in still images: the TextDiffuser paper on arXiv says diffusion models "currently struggle with rendering accurate and coherent text", with garbled or misspelled output as a common result. Video adds a second problem, because the text has to stay stable across every frame. That is why this effect restricts you to one or two single letters and does not accept full names.
How does this effect avoid scrambled letters?
It starts from an existing letter clip instead of drawing the letter from nothing. The effect keeps the black background, the fixed camera, the timing and the red glitch flicker of a top letter edit, then changes the letter to the one you tapped. Our own notes list the reasons: generating a letter from scratch often gives a scrambled or mirrored shape, and a repeated wall of text drifts into nonsense when regenerated from nothing. Editing an existing clip holds up better, but it is not a guarantee.
Has the effect been test-rendered, and will my letter be exact?
No. The effect has not been test-rendered, so we do not promise that every letter comes out exact. AI can still bend a stroke or change a letter. A failed, timed-out or blocked render returns its credits automatically, and you can check the letter in the finished clip before you post it. If a letter looks off, a second render is the practical fix.
Why does my skin and clothing turn red?
The look of the trend is red light. You appear behind the "I love you" lines, lit only by the glow, so skin and clothes take on a red cast. That is the style, not a fault. For the best likeness, use a front-facing photo with your head and shoulders in frame, even light, a plain top and a clean background, and keep your face uncovered. Sunglasses, hats and hands near the face make the likeness drift between frames.
Do I need permission to use someone else's photo?
Yes, get it. The effect takes one photo of you, and a photo of the two of you together also works, but the person in the picture should agree to its use. YouTube's privacy help page advises: "Get permission before filming other people or posting their personal information." The same habit applies to a video made from a friend's or partner's photo, particularly when you post it publicly or tag them. If in doubt, use your own picture.
How is this different from making a letter edit in CapCut?
A manual edit gives you full control and costs your time. You build the black background, the glow, the text wall and the timing by hand, then add a sound. The videosi route trades that control for speed: one photo, one letter tap, and a 15-second clip comes back in 9:16, 1:1 or 16:9. Use a manual editor if you want custom fonts, your own footage or frame-level timing. Use the effect page if you want you yourself behind the lines without building the timeline.
Does the video come with sound?
Yes. The finished video comes with the trend's original sound already on it. There is no lip sync, and you do not need to add music in another app. Letter edits use many sounds, and the top clip we studied is listed only as "original sound", so we do not name a track or an artist.
What does it cost?
The effect is 13 credits at the default render. A Single pack is $6.99 for 10 credits, a Trio Pack is $17.99 for 30 credits and a Duo Pack is $29 for 50 credits. Credits never expire, and a first video on an account never costs more than a 12-second one on the same tier, so the Single pack covers a first Standard 720p render. Check the page for the final price before you generate.
Ready to try it? Open the love name letter page, upload one photo, tap your letter and check the result before you post. The Seedance 2.0 page from ByteDance Seed describes a model that takes text, image, audio and video inputs together, which is the kind of reference-based setup this effect uses: Seedance 2.0 overview. Sources: TextDiffuser on arXiv and YouTube privacy guidelines.
