April 9, 2026
Slicing letters like salami
Charcuterie – Visual similarity Unicode explorer
Fans feast on a 'charcuterie board' of letters while arguing over a mysterious spotlight UI
TLDR: Charcuterie is a browser-based playground for exploring look‑alike letters and symbols, turning the giant Unicode set into a visual map. Commenters gushed over the name, sounds, and in-browser tech, while a vocal crew puzzled over the cryptic “spotlight” UI — sparking friendly debate about clarity versus vibes.
The internet just discovered Charcuterie, a playful visual explorer for all the world’s letters, symbols, and emojis, and the comments went full buffet. Fans swooned over the concept — “Everything runs in your browser,” cheered irickt — and the name got love too, with mplanchard calling it “very clever.” Behind the scenes, it uses modern image-matching AI to spot look‑alike characters, turning the massive Unicode set into a vibe-driven map you can wander. People fawned over the surprising sound effects (“natural thinking sounds”), and one user said it’s like paging through an alien alphabet primer.
But not everyone “got it.” The mysterious “spotlight” circle split the crowd: explorer-types happily clicked from letter to letter, while others, like siddboots, asked what the visual meta even means. evilelectron shouted “WOW!” at the experience; fortyseven went full cinema, comparing it to the UFO scribbles in Hangar 18. A few begged for clearer hints and tooltips; defenders said, relax — it’s a toy, a map, and a mood. Bonus points for privacy vibes: because it runs locally in the browser, no account, no upload. The dev says it’s still in active development and invites feedback — plus donations to keep the charcuterie board fresh.
Key Points
- •Charcuterie is a visual explorer for the Unicode character set.
- •It helps users discover visually related glyphs and learn about scripts, symbols, and shapes.
- •Visual similarity is computed by embedding rendered glyphs and comparing them in vector space.
- •The tool uses SigLIP 2 and DINOv2 models to generate glyph embeddings.
- •The project is under active development and welcomes feedback and donations.