Keep your Greek by reading the text.
63,963 entries dictionary entries · 1,895,013 inflected forms · self-host free, or Lector Cloud from $5 a month
Three years of Greek must not have a five-year half-life.
Morphology-aware lookup
1,895,013 inflected forms resolve to their lemma, so a tap answers the word in front of you and not a dictionary headword you must work back to. Lookups forgive accent variance, so τὸν finds τόν.
A real dictionary, on your device
63,963 entries entries and 93,508 English senses, from Ancient Greek Wiktionary through Kaikki, with per-word lemma data from MorphGNT for the Koine forms that Classical-leaning tables miss.
Corpus-frequency vocabulary
Frequency ranks come from the Greek New Testament itself through MorphGNT, with glosses from the Dodson lexicon. The 300 most frequent lemmas cover most of the running text, so early review buys the most reading.
Verse-aligned practice
7,329 verse segments of cloze practice, built from SBLGNT verses aligned to the World English Bible. Every practice sentence keeps its verse reference, so you can always check the context.
Polytonic typing, without the fight
Typed practice accepts a mark-stripped answer, because polytonic input needs a specialist keyboard. Type λογος and λόγος is accepted.
No audio, on principle
Koine pronunciation is reconstructed and disputed. Erasmian, Reuchlinian and period reconstructions all have a claim, so the speaker stays silent rather than teaching you one of them as fact.
Every number here comes from the pack page, which also names each source and its licence.
How far past the New Testament does it reach?
The dictionary reaches past the New Testament
The entries come from Ancient Greek Wiktionary, not from a New Testament word list. Homeric, Attic and Byzantine vocabulary is largely there, and the inflection index resolves forms from any period.
Reading has no corpus limit
Import an EPUB file, paste a passage, or load a page by URL. Homer, Thucydides, a Byzantine chronicle: the reader treats them all as text, and tap-to-define works the same way.
Frequency and practice are New Testament only
This is the real limit. Ranks come from the Greek NT, and every cloze sentence is a New Testament verse. For Homer the vocabulary order is wrong and the practice register is wrong.
What can change that
A Classical or Homeric frequency list gives a second spine, the way the Latin pack ranks on the Dickinson core vocabulary. A cloze bank outside the New Testament needs an aligned corpus. The Tatoeba Greek set holds about 1,800 sentences, which is too few.
Do you teach Homer, Attic, or Byzantine Greek? Come to the Discord and tell us which frequency list your students learn to. That decision needs a teacher, not a guess.
Latin ships too.
Biblical Hebrew needs readers, now.
What we ask, and what you get
- Tell us the text and the edition you read most.
- Tell us which lexicon your students already own.
- Read early builds, and say what got in the way.
- You get a direct line to the builder, and the pack you helped shape.
The boundaries, stated first.
It answers language questions
Grammar, morphology, vocabulary, and how a phrase works in context. It does not answer theological questions, and it stays denominationally neutral.
It brings no BDAG and no HALOT
Lexicon data is public domain or permissively licensed, and every source is named on the pack page. Keep your reference shelf. Lector does the reading and the retention, not the exegesis.
It is not an interlinear
An interlinear reads for you, which is exactly why the language fades. Lector shows you the text, answers the word you tap, and gives it back to you in review until it is yours.
Build progress is public on GitHub, and discussed as it happens on Discord.
The questions that matter most.
Is this a Bible-study or theology tool?
No. Lector is a language tool, and it stays denominationally neutral. It helps with grammar, vocabulary, morphology, and how a phrase works in context. It does not answer theological questions, and it never will.
What ships today, and what does not?
The Koine Greek pack ships now: 63,963 entries dictionary entries, 1,895,013 inflected forms, New Testament frequency ranks, and 7,329 verse segments of verse-aligned cloze practice. Biblical Hebrew does not ship yet. It is on the roadmap, and the cohort on this page is for Hebrew. The pack page lists every source and licence.
Does it cover Homer, Attic, and Byzantine Greek?
In part, and the split matters. The dictionary and the inflection index come from Ancient Greek Wiktionary. Vocabulary from Homer to the Byzantine period is largely covered, and you can read any text you import. The frequency ranks and every cloze sentence come from the Greek New Testament, so practice and word order suit the NT and not the Iliad. Teachers outside the NT are welcome in the Discord, because a second frequency spine is the next thing to build.
Which texts and lexica does it use?
Ancient Greek Wiktionary entries through Kaikki, with lemma data from MorphGNT. Glosses come from the Dodson Greek-English lexicon. Cloze sentences come from SBLGNT, aligned to the World English Bible. All of it is public domain or permissively licensed. There is no bundled BDAG and no HALOT, so keep the reference shelf you already own.
Why does the speaker button stay silent?
Because Koine pronunciation is reconstructed and disputed. Erasmian, Reuchlinian and period reconstructions each have a case, and synthesized audio can only teach one of them as settled fact. The same rule applies to Latin, where classical and ecclesiastical pronunciation disagree.
What about Hebrew?
Second, and deliberately. Pointed Hebrew is harder engineering than Greek, though less so now. The Arabic pack shipped as the first right-to-left pack, so the reader handles right-to-left text today. Biblical Hebrew is on the roadmap, and the people who write in now shape its priority.
Is Latin ready too?
Yes. The Latin pack ships with 833,680 entries entries and 965,021 inflected forms. Keys strip editorial macrons, so amāre and amare are one key, and the u/v and i/j spellings that editions mix both resolve. Frequency ranks come from the Dickinson College Commentaries core vocabulary. See the Latin guide.
Do I need to be a scholar?
No. Morphology help on tap means you can read above your parsing level while it comes back. This suits a pastor who prepares every week, a student in the middle of coursework, and anybody whose hard-won Greek slips away.
How is this different from Logos or an interlinear?
A reference platform is excellent at exegesis lookup, and an interlinear does the reading for you, which is why the language fades. Lector is a loop. Read the text, tap what you do not know, and keep what matters. Review it until it is yours, in the app and in Anki.
Can I use it with no internet connection?
Yes. The reader, the on-device dictionary and cloze practice all run with the network off. Read offline and remote use for the detail.
What does it cost?
Self-hosting is free forever, and it is open source, with every language pack included. There is no per-language price and no scholar tier. Lector Cloud starts at $5 a month if you prefer not to run a server.
Do not let it fade.
The Greek pack is ready. Import a chapter tonight, tap what you have lost, and let review give it back.
Reading the Hebrew Bible, or teaching outside the New Testament? Write to us at [email protected].