Anton is the heavy face Snap captions use
/fonts/anton is Heavy & Bold — Snap's pop effect is built on it. Loading Anton as a static subtitle skips the scale-and-recolor job the timed group is for.
/fonts/anton is Heavy & Bold — Snap's pop effect is built on it. Loading Anton as a static subtitle skips the scale-and-recolor job the timed group is for.
Anton is a caption and overlay face in the Versely font registry. It is not an Ideogram prompt. It is not a generate. The Snap family uses fontId anton because the face has the mass to survive a scale-and-recolor on every spoken word. Putting Anton on a static lower-third gives you the weight without the job.
Anton is a caption face, not a stills model
The font record is specific: id anton, name Anton, category display, preview string Heavy & Bold. In the catalog copy it is heavier than Bebas, tighter than Impact — maximum mass per character. The honest limit is also specific: above about five words per line it becomes a wall. Use it where words land one or two at a time.
That limit is why Snap exists. Word-timed captions feed Anton a token, not a paragraph. A static subtitle block in the same face is a brick of caps. Anton is also in the uppercase set the renderer uses for caption previews — the Snap line reads as shouted mass, not as book text.
Do not send "Anton, Heavy & Bold" to an image model and hope the letters come back editable. Generated type in a poster is a misspelling you regenerate to fix. Anton is the face the caption renderer already ships. Burn it. Do not paint it.
The rest of the registry lives on the fonts index. Bebas is the Headline family. Anton is Snap. Switching those two because both look "loud" is how a word-pop hook becomes a title card, or a chyron becomes a punch that never punches.
Snap is timed pop, not a static subtitle
Snap is a caption group. Kind is timed, not subtitle. Effect is pop. Position is bottom. fontId is anton on every variant. White fill, black outline, outline width 1. Each spoken word scales up and recolours — punchy and energetic. That is the group description, not a vibe.
The base preset is Snap (word_pop): accent #FF4785. Siblings keep the same face, the same pop, the same bottom slot, and change only the accent:
- Snap Gold —
#FFD700 - Snap Cyan —
#22D3EE - Snap Lime —
#A3E635 - Snap Mono —
#FFFFFF(no hue jump; scale only)
Mono is the quietest Snap: the word still scales, it does not recolor. That is still timed. It is not a subtitle family wearing Anton.
Seven of the nine caption families are subtitle-style: the whole line appears and updates together. Snap and Reveal are the two word-timed families. Reveal builds words that stay. Snap punches one token at a time. Against a slow, considered read, Snap looks frantic. Against a fast hook, a static Anton line looks like a transcript of an interrupt, which is a contradiction.
Burn-in is not a generate. You apply Snap to a take that already has speech (or a voiceover you already recorded). The family times to that audio. It does not invent the line, and it does not invent the face.
Loading Anton without Snap skips the job
Three different objects get confused because they can all print Anton-looking caps:
Snap group. Word-timed, pop effect, Anton, bottom. Requires a clock per word. That is the job this page is for.
A static subtitle family with Anton forced as the face, or a persistent overlay in Anton for the whole clip. You get Heavy & Bold. You do not get scale-and-recolor. The token never pops. Fine for a badge. Wrong for a Snap hook.
Timed text overlays. Add timed text overlays to a video burns lines you wrote at start_sec / end_sec. Kind is authored overlay, not Snap. Nothing is transcribed. Font can be Anton if you pick it. The renderer still will not run the pop effect on each spoken word, because there is no spoken-word clock on that path. You are sequencing cards, not punching tokens.
If the words are speech, add captions to a video and pick the Snap group. If the words are a hook nobody said, overlay or timed overlay. If you only wanted the face on a mute product plate, Anton as overlay is allowed — just do not call it Snap.
Pick the accent last; keep the face
Once Snap is the family, the variant is brand colour, not a new typeface. Gold, cyan, lime, or the pink base. Mono when you want the timing without the flash. Do not hop to Headline/Bebas because a title-card mock looked cleaner in a still. Headline is subtitle kind, top-weighted, a different job.
Do not set Snap on a take you are about to recut. Word clocks attach to this audio. A new generate is a new clock. Picture lock, voice lock, then burn.
Anton on a five-word line is the wall the caveat named. Rewrite the line. Do not shrink the face until it is a different font. The Snap group is built for one or two words on screen at a time. That is the heavy-face contract.
FAQ
Can I use Anton without the Snap pop effect?
Yes, as a caption/overlay face. You get Heavy & Bold without scale-and-recolor. That is a static job. It is not the Snap group.
Is Snap a subtitle preset with Anton swapped in?
No. Snap's kind is timed. Subtitle families show the whole line together. Snap pops each spoken word. Anton is the face that job runs on.
Does add-timed-text-overlays apply Snap's pop?
No. Timed overlays are authored lines on a clock you type. They do not transcribe, and they do not run the Snap pop. For spoken-word Snap, caption the take and pick the group.
Why not generate the words in Anton inside the video model?
Because then you cannot change a spelling, a price, or a language without a new generate, and you never get word timing. Anton is a renderer face. Snap is a timed burn. Neither is a prompt.