- Starting price
- Free; voluntary support planned
- Free tier
- Yes
- Platforms
- Web · iPhone · Android · Apple-silicon Mac
- Developer
- Venante Ducrepin
- Launched
- 2026
- Updated
- Aug 9, 2026
The verdict
ScOAR is better scoped than a general chatbot: describe a circumstance, receive several candidate passages, verify them and keep reflecting. Its FAQ openly admits AI can make mistakes, documents its canon and translation choices, and tells readers not to replace church or wise counsel. Yet the same FAQ and App Store copy say guest searches stay on-device while the privacy policy correctly says query text passes through Vercel to Anthropic. Local history is not local processing. Until that wording is fixed, users should assume every submitted struggle is disclosed to external infrastructure and omit details they would not send to an AI provider.
Try ScOAR ↗Opens scoarapp.org
ScOAR stands for Scriptural Orientation, Application & Reflection. It is a free AI-assisted Bible guidance tool created by Venante Ducrepin. A user describes stress, grief, a relationship, doubt or a decision in plain language. Anthropic’s Claude produces a short summary, three to five Bible references, an explanation and practical action for each, followed by a “Moving Forward” reflection. The user can request more verses, copy or share results, save favorites locally, or create an optional account and place selected query/response pairs in a cloud-based Private Vault.
That tight workflow is its best design decision. A general chatbot invites almost any prompt and can produce a persuasive essay with few anchors. ScOAR constrains the output toward Scripture candidates and actions. The official FAQ identifies the 66-book canon and a broadly evangelical calibration, admits that interpretation can be wrong, and encourages verification with one’s own Bible and a trusted pastor or spiritual advisor. Those disclosures help readers understand that the software is not the biblical text and that its doctrinal frame is not universal or neutral.
ScOAR currently supports six interface/response languages: English, Spanish, French, Portuguese, Chinese and Korean. The FAQ says English responses tend to be theologically more precise because the underlying model was trained predominantly on English material; other-language guidance is AI-generated, not human-translated. Public-domain Bible text is used to avoid licensing costs. English currently documents WEB, ASV and KJV, with French LSG; the FAQ’s broader list also names RVR1960, ARC, CUV and KRV, while another explanation says verse-text retrieval is currently limited by API availability in some languages. Users should verify what the live result actually quotes rather than relying on an aspirational catalog.
The product says it will remain completely free, with no ads, subscription or premium tier, supported through voluntary donations and sponsorships. Apple lists a 5.0 US score from one rating, version 2.0 from 29 May 2026, a small 5.7 MB download and iOS 15 or later. The official FAQ still says the native stores are “coming soon” in one platform section even though both listings are live, revealing documentation lag. This review is based on public materials and stores. We did not submit a sensitive query, audit model prompts, inspect requests, test vault access, compare multilingual outputs or independently red-team crisis responses.
✓ The good
- Narrow Scripture-first workflow - results are organized around candidate passages rather than an open-ended chatbot monologue
- No account required - core search, local history and favorites can be used without giving ScOAR an account identity
- Completely free and ad-free - there is no subscription, premium tier or advertising network in the stated model
- Specific AI disclosure - the FAQ names Anthropic Claude and the privacy policy lists Vercel, Supabase, bolls.life and Cloudflare
- Meaningful limitations published - ScOAR tells readers to verify verses and interpretations and not replace prayer, church or wise counsel
- Six response languages - useful reach, accompanied by an unusually candid warning that non-English nuance may vary
- Public-domain Scripture strategy - keeps access free and identifies at least several supported editions
- Immediate in-app account deletion - the privacy policy says account and Vault entries can be removed from the app without waiting for email
✗ Watch out
- False local-query impression - the FAQ says guest queries never reach servers, but query content must pass through Vercel to Anthropic
- AI can invent or misapply Scripture - a confident explanation and action step may be wrong even when the reference exists
- No cited commentary provenance - Claude’s learned theological framing is not a source list a reader can inspect or challenge
- Vault is not documented as end-to-end encrypted - Supabase encryption and row-level security do not prove operators can never access plaintext
- Translation documentation conflicts - the FAQ mixes a broad edition list with API limitations and AI-generated non-English guidance
- Organizational identity is thin - “ScOAR App Team” is named without a legal entity, physical address or fuller controller details
- Age and crisis stakes - a 13+ audience may bring self-harm, abuse or medical concerns to a system unable to provide emergency care
- Minimal independent evidence - one US Apple rating and a launch-day version cannot establish dependable accuracy or security
Best for
- Adults and older teens seeking possible Scripture starting points for an ordinary-life question
- New Bible readers who need help translating a feeling into passages to examine
- People willing to verify every quotation, context and application outside the AI result
- Users who want a focused free tool rather than a general-purpose chatbot
- Church leaders evaluating AI literacy with non-sensitive example prompts
Avoid if
- The prompt contains names, health details, abuse disclosures, legal facts or information another person did not consent to share
- You need confidential pastoral care, emergency response or licensed professional advice
- Your community’s canon or doctrinal standards differ materially from ScOAR’s stated calibration
- You require human-reviewed translation and interpretation in every supported language
- You would treat a fluent AI response as a personalized word from God or a final biblical interpretation
What ScOAR is
ScOAR is a structured front end around a language model and Bible-verse lookup services. The user supplies at least a short description; Claude selects references and generates summaries, explanations, action steps and reflection. Public-domain verse text is retrieved for supported editions through bolls.life, rather than trusting the language model to reproduce every quotation. The result can be saved, shared or expanded with three additional suggestions.
It is not a search engine that retrieves live articles, a human-curated counseling database or a commentary whose source trail is visible. The FAQ says Claude reasons from learned material rather than browsing. Training exposure does not allow a reader to identify which theologian, commentary or tradition shaped a sentence. ScOAR’s own prompt and calibration influence output, but cannot guarantee a stable interpretation.
Why constrained AI is safer than open chat—and still not safe by default
ScOAR limits the conversational surface. Instead of inviting a long back-and-forth in which the model imitates a pastor, it produces a recognizable package: candidate verses, reasons, one concrete step and further reflection. Multiple verses make comparison possible, and external verse retrieval can reduce fabricated quotation. Explicit warnings remind the reader to open a Bible. This is a healthier product shape than a chatbot that quietly turns probability into spiritual authority.
The model still decides relevance and writes interpretation. It may select a real verse whose speaker, covenant setting, genre or surrounding argument makes the application misleading. It may flatten lament into optimism, prescribe reconciliation where safety requires distance or give a generic action to a complex illness. The UI’s polish can increase trust beyond accuracy. Users should separate four questions: Is the quotation exact? Is the reference understood in context? Is this application warranted? Is this step safe for this person?
Three to five verses with actions: useful scaffolding that needs verification
A standard result includes a brief orientation, three to five passages, an explanation of why each might apply, one action for today and a “Moving Forward” section. Simpler requests receive fewer verses; more complex ones may receive five, and a control can request three more. The structure encourages movement between reading and response rather than presenting a single slogan. It can also teach new readers how to articulate a possible connection between text and life.
Verify in layers. Confirm the words in the named translation, because a reference can be right while a generated paraphrase is wrong. Read the whole literary unit. Compare another translation where phrasing carries the argument. Ask whether the action follows from the passage or merely sounds wholesome. For disputed doctrine or life-altering decisions, consult qualified people who know both Scripture and the situation. Do not repeatedly regenerate until the software gives the answer already desired.
Local history and the Private Vault: two storage modes, one processing path
Guests can keep history and favorites in local browser or app storage. Registered users can “lock” selected query/response pairs into a Supabase database for cross-device access. The policy says account records are encrypted at rest and in transit and protected with row-level security so another ordinary account cannot read them. In-app deletion is described as immediate; email deletion is promised within 30 days.
Neither mode keeps query text entirely on the device. To generate guidance, ScOAR forwards it through a Vercel serverless proxy to Anthropic. The privacy policy says ScOAR does not log that proxy content and does not use it for model training, but Anthropic’s own terms govern its processing. The FAQ’s statement that guest queries are “never sent to our servers” is therefore materially misleading even if it means they are not retained in ScOAR history. Assume transmission, minimize the prompt and avoid third-party secrets.
Six languages and public-domain Bibles: access with uneven assurance
The interface and AI guidance support English, Spanish, French, Portuguese, Chinese and Korean. ScOAR candidly states that English tends to be more theologically precise and that other-language explanations are generated directly rather than reviewed human translations. Public-domain editions avoid subscription and licensing barriers. English options include the World English Bible, American Standard Version and King James Version, while French verse text is retrieved through the verse API.
The documentation needs consolidation. One FAQ answer lists RVR1960, ARC, CUV and KRV as current translation choices, while another says direct verse text for Spanish, Portuguese, Chinese and Korean is unavailable because of API limitations. Some named editions may also involve jurisdiction-specific copyright questions. A user should identify whether the screen shows verified Bible text, a reference only or AI-generated guidance; then compare the result with an authorized Bible in that language. Language access is valuable, but labeling must be exact.
Pricing
Guest use
Free
Search, response generation, local history and favorites require no account and no payment. History can remain in browser/device storage, but every query is still transmitted through ScOAR’s Vercel API proxy to Anthropic for processing. “Guest” means no ScOAR account, not offline or undisclosed use.
Optional account
Free
An email, username and password enable Supabase authentication, cross-device access, password recovery and the Private Vault. The policy says no paid entitlement is attached. Use it only when cloud convenience justifies an identity-linked record.
Private Vault
Included
Selected query/response pairs can be moved from local history into account-linked cloud storage protected by row-level security and encryption at rest/in transit. “Private” should be understood as access-controlled cloud storage, not proven client-side end-to-end encryption.
Voluntary support
Donation or sponsorship
The FAQ says donations and sponsors sustain the ministry, though no paid benefits are described. Support does not make an AI answer more authoritative and should not be confused with purchasing professional advice or a service-level guarantee.
ScOAR states a strong access commitment: completely free, always, with no paywall, premium tier, subscription or advertising. Every documented feature—including accounts and the Vault—is presented as included.
That model avoids monetizing a vulnerable moment. A person does not have to subscribe after describing grief or pay to recover saved guidance. The policy also says there are no ad networks, social trackers or sale of personal information.
AI inference, serverless hosting, authentication and databases still cost money. Donation support can sustain access, but the site does not publish a service guarantee, financial report or closure/export plan. Users should not let a free cloud vault become the only copy of meaningful reflections.
The hidden cost is verification time. A zero-dollar result is not free if a confident error misdirects someone. Budget attention for reading context and human counsel. ScOAR can reduce discovery friction; it cannot outsource responsibility.
Where ScOAR falls behind
The guest-query contradiction must be corrected. The FAQ says guest queries, history and favorites stay only on the device, that ScOAR never sees them and that they are never sent to its servers. The privacy policy says every query is forwarded through a Vercel API proxy to Anthropic. Only history and favorites are local. Marketing should say “not retained in your ScOAR history,” not “never sent.”
The Private Vault copy overstates technical assurance. FAQ language says no one, “not even us,” can see entries, while the policy describes conventional server-side encryption and Supabase row-level security. Those controls restrict access but do not establish user-held keys or end-to-end encryption. The developer should publish a threat model and use precise access language.
Model provenance cannot support the FAQ’s “well-read biblical scholar” analogy. Claude is not a person who can name books studied, distinguish memory from inference or accept pastoral accountability. The claim that training included the Bible, commentaries, theology and scholarship does not provide citations for a particular response.
Documentation already lags a new product. The FAQ says App Store and Google Play availability is coming soon even though the native listings launched in May 2026. The translation sections also conflict. A versioned documentation owner and public change date would improve trust.
Operator details are insufficient for sensitive processing. The privacy policy names the “ScOAR App Team” but not a registered entity, controller address, hosting region, Anthropic data-retention setting or backup behavior. An email channel is helpful, not a substitute for a complete identity and data map.
Crisis boundaries require stronger interface controls. A written warning to seek pastoral or professional help is appropriate, but users aged 13+ may enter self-harm, abuse, violence or medical emergencies. The product should publish localized crisis behavior, escalation limitations, evaluation results and safe failure rules without implying its model can reliably detect every emergency.
ScOAR vs. Bible Answers AI vs. ChatGPT
Bible Answers AI is another specialized question-to-Scripture product and may provide a more search-like knowledge experience. ScOAR is specifically structured around a personal situation, several passages, actions and reflection. Compare their source transparency, model disclosure, privacy and correction paths rather than choosing the one that sounds more certain. A cited human-authored answer can be easier to audit than an elegant generated one.
ChatGPT is much broader: it can compare translations, explain historical background and follow a complex conversation, but that flexibility also lets users frame it as an oracle, counselor or simulated pastor. ScOAR’s constrained response and recurring verification warning are preferable for a simple starting point. Neither product turns probabilistic text generation into revelation or professional care.
A Bible concordance or trusted human leader is slower but offers different accountability. A concordance shows lexical occurrences without claiming personal relevance; a pastor can ask questions, know context and remain responsible for counsel. ScOAR is best in front of those resources: use it to surface possibilities, then leave the product to test them. It is weakest when it becomes the only voice in a private crisis.
The bottom line
ScOAR has a sensible shape for an AI Bible tool. It names its provider, narrows the task, retrieves public-domain verses, gives multiple candidates, publishes doctrinal and language boundaries, and repeatedly tells readers to verify. Those choices deserve credit. The privacy contradiction is not a footnote, because the very text people submit may contain their most vulnerable circumstances. Guest history can be local while query processing is remote; ScOAR must say both. Use short, anonymous prompts, verify the entire result and keep the app outside emergency, confidential and life-altering decisions. With corrected privacy language, consistent translation documentation and stronger vault claims, it could become a responsible discovery aid rather than another opaque spiritual chatbot.
Alternatives to ScOAR
Bible Answers AIA focused AI Bible question tool to compare on citations, privacy and interpretive transparency.
FaithTime AIAn AI-assisted devotional product with a broader habit and content experience.
Kaleo AIA Christian AI assistant with a different feature scope and similar need for verification.
OpenBibleA topical Bible reference site that surfaces community-voted passages without personalized generated counsel.A trusted pastor and full BibleSlower and less available on demand, but able to know context, ask questions and accept relational accountability.Frequently asked questions
What does ScOAR do?
A user describes a life situation in plain language. Anthropic Claude generates a summary, three to five relevant Bible references, explanations, practical action steps and a “Moving Forward” reflection. Supported verse text is retrieved through a Bible API. Results can be expanded, copied, shared, saved locally or placed in an account-linked Vault. It is a discovery and reflection aid, not an authoritative interpretation or counselor.
Is ScOAR free?
Yes. The official FAQ says every feature is free with no advertising, subscription, paywall or premium tier. Optional donations and sponsorships are intended to sustain it. Apple and Google list free downloads. Free access does not guarantee permanent service or accuracy, so users should preserve important notes and verify every generated claim.
Do guest ScOAR queries stay on my device?
History and favorites can stay on the device, but the query itself does not. ScOAR’s privacy policy says submitted text passes through a Vercel API proxy to Anthropic Claude to generate the response. It says ScOAR does not log or retain that content unless a user saves it to the Vault. The FAQ and store wording that queries “stay on your device” is therefore incomplete and should be corrected.
Is ScOAR’s Private Vault end-to-end encrypted?
The public policy does not establish end-to-end encryption or user-held keys. It says Vault records are encrypted at rest and in transit in Supabase and protected by row-level security so only the authenticated account can normally query them. Those are useful controls, but they do not justify a literal guarantee that no operator, processor, backup system or compelled party could ever access plaintext.
Which Bible translations and languages does ScOAR support?
Responses support English, Spanish, French, Portuguese, Chinese and Korean. The FAQ names WEB, ASV and KJV for English and several public-domain or older editions for other languages, but another section says verified verse-text retrieval is currently available only for English and French while other languages receive AI guidance. Because the documentation conflicts, inspect the live label and verify every verse in an authorized Bible.
Can ScOAR give wrong Bible advice?
Yes, and the developer acknowledges that AI can make mistakes. Claude may fabricate wording, select a real passage with the wrong context, import an unstated theological assumption or give an unsafe generic action. Confirm the quotation, read the surrounding section, compare reliable resources and consult a trusted leader. Never keep regenerating answers until one validates a desired decision.
Can I use ScOAR for a crisis or mental-health emergency?
Do not rely on it for crisis detection, emergency instructions, diagnosis, abuse response, legal advice or treatment. Its 13+ audience and sensitive-topic framing do not turn the model into a professional service. If someone may be in danger, contact local emergency or crisis resources and a trusted adult or qualified professional. You can pray and read Scripture while also taking immediate responsible action.
Sources & further reading
More AI Bible Apps
Faith: Scripture MeditationFaith turns a Bible passage or personal concern into a narrated meditation with Scripture, breathing and music—an inviting format whose undisclosed AI data path deserves more scrutiny than its local-storage promise suggests.
GamalielGamaliel pairs a Bible reader with free AI questions, source links, audience profiles and publicly inspectable theological prompts—an unusually transparent experiment whose deployed model, retention and store disclosures still require independent caution.
Agentic JesusAgentic Jesus is a proactive Claude-powered faith companion with Bible chat, adaptive devotionals, prayer memory, voice and follow-ups—and its unusually clear data map also shows why intimate spiritual AI needs firm human boundaries.
LumenologyLumenology combines a multi-translation Bible with cited AI research, interlinear data, historical commentary and passage tools in a clean workspace—but citations reduce risk rather than removing it, and every important claim still needs verification.
My BibleVerseMy BibleVerse turns an emotion or written concern into a suggested Bible passage, reflection and AI-written prayer, then saves the moment into a daily habit—but its costly weekly plan, intimate data processing and confident personalization claims require discernment.
Polaris BiblePolaris is an unusually transparent conservative evangelical AI Bible companion with a capable reader and low pricing—valuable for its intended audience precisely because its lens is explicit, not because guardrails make AI infallible.