Show HN: I'm a pastor/dev and built a 200M token generative Bible
anselm-project.comAre they matching against a simple regex?
Inconsistent example format:
```
Enter any Bible reference - single verses, ranges, or entire chapters. Examples: "Matthew 5:3-12", "Psalm 23", "1 Corinthians 13"
Invalid reference format (e.g., John 3, 1 Cor 13:4-7, Gen 1-2). Book name only is not allowed.
```
I also added about 20 credits for you, if you feel up to trying the engine.
``` BIBLE_ABBREVIATIONS = { 'Gen': 'Genesis', 'Gn': 'Genesis', 'Exod': 'Exodus', 'Ex': 'Exodus', 'Exo': 'Exodus', 'Lev': 'Leviticus', 'Lv': 'Leviticus', 'Num': 'Numbers', 'Deut': 'Deuteronomy', 'Dt': 'Deuteronomy', 'Josh': 'Joshua', 'Jos': 'Joshua', 'Judg': 'Judges', 'Jdg': 'Judges', '1 Sam': '1 Samuel', '1Sam': '1 Samuel', '1st Sam': '1 Samuel', '1S': '1 Samuel', '1 S': '1 Samuel', '2 Sam': '2 Samuel', '2Sam': '2 Samuel', '2nd Sam': '2 Samuel', '2S': '2 Samuel', '2 S': '2 Samuel', '1 Kgs': '1 Kings', '1Kings': '1 Kings', '1st Kings': '1 Kings', '1K': '1 Kings', '1 K': '1 Kings', '2 Kgs': '2 Kings', '2Kings': '2 Kings', '2nd Kings': '2 Kings', '2K': '2 Kings', '2 K': '2 Kings', '1 Chr': '1 Chronicles', '1Chron': '1 Chronicles', '1st Chronicles': '1 Chronicles', '2 Chr': '2 Chronicles', '2Chron': '2 Chronicles', '2nd Chronicles': '2 Chronicles', 'Neh': 'Nehemiah', 'Esth': 'Esther', 'Est': 'Esther', 'Ps': 'Psalms', 'Psa': 'Psalms', 'Prov': 'Proverbs', 'Pro': 'Proverbs', 'Eccl': 'Ecclesiastes', 'Ecc': 'Ecclesiastes', 'Song': 'Song of Solomon', 'SOS': 'Song of Solomon', 'Cant': 'Song of Solomon', 'Isa': 'Isaiah', 'Jer': 'Jeremiah', 'Lam': 'Lamentations', 'Ezek': 'Ezekiel', 'Eze': 'Ezekiel', 'Dan': 'Daniel', 'Hos': 'Hosea', 'Obad': 'Obadiah', 'Ob': 'Obadiah', 'Mic': 'Micah', 'Nah': 'Nahum', 'Hab': 'Habakkuk', 'Zeph': 'Zephaniah', 'Zep': 'Zephaniah', 'Hag': 'Haggai', 'Zech': 'Zechariah', 'Zec': 'Zechariah', 'Mal': 'Malachi', 'Matt': 'Matthew', 'Mt': 'Matthew', 'Mk': 'Mark', 'Mrk': 'Mark', 'Lk': 'Luke', 'Luk': 'Luke', 'Jn': 'John', 'Jhn': 'John', 'Rom': 'Romans', 'Rm': 'Romans', '1 Cor': '1 Corinthians', '1Cor': '1 Corinthians', '1st Corinthians': '1 Corinthians', '2 Cor': '2 Corinthians', '2Cor': '2 Corinthians', '2nd Corinthians': '2 Corinthians', 'Gal': 'Galatians', 'Eph': 'Ephesians', 'Phil': 'Philippians', 'Php': 'Philippians', 'Col': 'Colossians', '1 Thess': '1 Thessalonians', '1Thess': '1 Thessalonians', '1st Thessalonians': '1 Thessalonians', '2 Thess': '2 Thessalonians', '2Thess': '2 Thessalonians', '2nd Thessalonians': '2 Thessalonians', '1 Tim': '1 Timothy', '1Tim': '1 Timothy', '1st Timothy': '1 Timothy', '2 Tim': '2 Timothy', '2Tim': '2 Timothy', '2nd Timothy': '2 Timothy', 'Phlm': 'Philemon', 'Phm': 'Philemon', 'Heb': 'Hebrews', 'Jas': 'James', 'Jms': 'James', '1 Pet': '1 Peter', '1Pet': '1 Peter', '1st Peter': '1 Peter', '2 Pet': '2 Peter', '2Pet': '2 Peter', '2nd Peter': '2 Peter', '1 Jn': '1 John', '1John': '1 John', '1st John': '1 John', '2 Jn': '2 John', '2John': '2 John', '2nd John': '2 John', '3 Jn': '3 John', '3John': '3 John', '3rd John': '3 John', 'Rev': 'Revelation', 'Rv': 'Revelation' } ```
Fails with:
```
Matthew 5:31-32; Matthew 19:3-9; Mark 10:2-12; Luke 16:18; 1 Corinthians 7:10-16; 1 Corinthians 7:39; Romans 7:2-3
```
```
"Matthew 5:31–32", "Matthew 19:3–9", "Mark 10:2–12", "Luke 16:18", "1 Corinthians 7:10–16", "1 Corinthians 7:39", "Romans 7:2–3"
```
```
Matt 5:31-32, 19:3-9; Mk 10:2-12; Lk 16:18; 1 Cor 7:10-16, 7:39; Rom 7:2-3
```
```
Matt 5.31-32, 19.3-9; Mk 10.2-12; Lk 16.18; 1 Cor 7.10-16, 7.39; Rom 7.2-3
```
```
Matthew 5:31-32 Matthew 19:3-9 Mark 10:2-12 Luke 16:18 1 Corinthians 7:10-16 1 Corinthians 7:39 Romans 7:2-3
```
Maybe the validator isn't splitting the string on separators before parsing, or choke on leading numerals or spaces, or normalization to ascii?
It's probably my desire as a expository preacher to only handle a passage in its immediate context when I can help it.
But it still doesn't seem to work.
I tried `Matthew 19:3–9` and `"Matthew 19:3–9"` and still warns of invalid formatting.
I will think about what would go into a topical studies engine.
Eventually, it will be a back and forth discussion between AIs batting a question around by a mixture of experts system with a summary report provided on theological questions. Topical would obviously be included in that.
I'll work immediately on that. Really sorry - but that's the exact thing I needed. Thanks for finding that. I'll try and add a few more credits so you can test more and find more problems.
Also, the Synod "works" but isn't quite ready for primetime... or even an early morning jog.
I'm a $43k-level power-user of Logos Bible Software, but I was still spending 10-20 hours a week on the exegetical work—collating word studies, parsing Greek/Hebrew, and comparing parallels for my sermons. I built the tool I wished I had.
The result is the Anselm Project. It's two parts:
The Engine (Free): A 200M+ token generative database of the entire 66-book Bible. It uses LLMs to generate a full scholarly apparatus (morphology, parallels, and in-progress text-crit) for every verse. You can browse this for free. There's no login needed for this part.
The SaaS (Paid): A report generator on top. It synthesizes all that data into 40-page sermon prep reports or 60-page academic briefs. I'm a solo dev, not a marketer. I'd also love to retire someday. Ministry doesn't pay much.
The tech: The 200M+ token dataset was generated via API calls in a massive batch process. My cost per academic report is about $0.07. I'm a solo dev, not a marketer.
The free generative Bible is open for browsing. I'd love your feedback on the tech, the model's output, and the (likely buggy) UI. Happy to answer any questions about the data generation, costs, or the tech stack.
Thanks.