atlas.dict is a fast, offline terminal dictionary with a delicate design system. Part of the Atlas Suite, it ships embedded language-pair data and looks up rich entries (parts of speech, senses, examples, synonyms) in milliseconds.
- 📚 Offline-first: Dictionary data is embedded in the binary; no network calls.
- ⚡ Fast: O(log n) binary-search index over a packed format; per-entry zstd decoding.
- 🎨 Delicate design: Restrained lipgloss palette — airy, readable, no boxy chrome.
- 🌐 Multi-pair: Built around language pairs; new pairs add
.dictfiles. - 🖥️ CLI + TUI: One-shot lookups for the terminal, browse mode for exploration.
$ atlas.dict -f en -t tr -q "eminence"
eminence /ˈɛmɪnəns/ · noun
saygınlık · yücelik$ atlas.dict -f en -t tr -q "eminence" --full$ atlas.dict -f en -t tr -q "eminence" --short --plain
saygınlık$ atlas.dict --pairs$ atlas.dict| Key | Action |
|---|---|
| type | live filter |
| ↑/↓, k/j | navigate results |
| Enter | open entry |
| Tab | swap from↔to |
| F | toggle compact / full |
| Ctrl+P | language-pair picker |
| Esc / q | quit |
gobake build # cross-compile all targets
gobake data # regenerate embedded .dict filesBinaries land in build/.
The embedded dictionary data is sourced from
FreeDict — specifically the
eng-tur bundle
(36,589 headwords). A small set of hand-curated entries lives alongside
the FreeDict data in
scripts/build-dict/sources/ and takes
priority on conflict. The reverse tr-en pair is auto-synthesized
from the forward direction at build time.
To regenerate the embedded .dict files (e.g. after upgrading FreeDict):
gobake data # downloads + parses + writes internal/data/*.dict- atlas.dict source code: MIT — see LICENSE.
- FreeDict dictionary data: GNU GPL v2 or later (gnu.org/licenses/gpl-2.0.html). Copyright (C) 1999–2017 by the contributors listed in the original FreeDict source.
Because the released binary embeds FreeDict's GPL data, the combined binary distribution is subject to the GPL v2+. The atlas.dict source remains MIT and can be used in any context where you supply your own dictionary data.
