canonical: https://jentic.com/apis/itsbaba.com/itsbaba # Itsbaba HebrewCore API The HebrewCore API translates text between Hebrew, English, Russian, and other languages and returns RTL and LTR metadata for both sides. Protected translation keeps prices, quantities, dates, codes, product names, and interface variables exactly as written and verifies every sentence, with an account-level list of protected terms such as drug or brand names. Beyond translation it detects language in code-switched input, transliterates between Hebrew and Latin script, adds or strips nikud vowel points, and wraps mixed Hebrew and English lines for display. A Clinical API adds verified clinical English-to-Hebrew translation, streaming over Server-Sent Events or a WebSocket, Hebrew-to-English record queries, and translation verification. ## For AI agents Translate to and from Hebrew with protected numbers, dates, and terms, detect language, transliterate, add nikud, wrap bidirectional text for display, and run verified clinical translation. ## Scope Does not handle speech-to-text, document file conversion, or translation project management. Use for Hebrew-centred text translation and language processing only. ## Capabilities - Translate text or a batch of strings into Hebrew, English, or Russian with a quality tier of fast, balanced, or smartest - Protect prices, dates, codes, and listed protected_terms so they survive translation exactly as written - Keep an account list of protected terms such as drug and brand names, added by upsert and removed by id - Detect the language of code-switched Hebrew and English input and analyse its text direction - Transliterate he-to-latin or latin-to-he, and add or strip nikud vowel points for display and speech - Wrap a mixed Hebrew and English line for display as html, isolate, or marks output - Translate clinical English into verified Hebrew for a clinician or patient audience, and verify an existing translation ## Use cases ### Agent-run Hebrew localisation An AI agent localising an app's interface strings into Hebrew needs translations that leave placeholders, prices, and product names untouched. Through Jentic the agent searches for 'translate text into Hebrew', loads the HebrewCore batch operation, and sends the strings with target_lang he and protect enabled, with the bearer key injected at call time. The response carries direction metadata, so the agent can flag strings that need RTL layout. Example prompt: Call POST `/v1/translate/batch` with items Welcome to your account and Your order total is $49.99, target_lang he, and protect true, then return the Hebrew strings ### Clinical content for Hebrew-speaking patients A health product generating English discharge notes needs Hebrew versions it can rely on. The Clinical API translates clinical English into verified Hebrew with an audience of clinician or patient and a patient_gender option, keeps protected and redacted terms, and the verify operation checks a Hebrew translation against its English source with strict or standard verification. Example prompt: Call POST `/v1/medical/translate` with text Take one tablet twice daily after meals and audience patient, then call POST `/v1/medical/verify` with the English source and the returned Hebrew translation ### Bidirectional text rendering A messaging app showing mixed Hebrew and English lines needs them to render in the right order. The wrap operation returns a display-ready line as html, isolate, or marks output with a base direction of auto, ltr, or rtl, and the detect operation identifies code-switched input before it is displayed. Example prompt: Call POST `/v1/wrap` with a mixed Hebrew and English line, format html, and base auto, and return the wrapped output ### Hebrew for speech and search A voice assistant reading Hebrew aloud needs vowel points to pronounce words correctly, and a search index needs Latin spellings of Hebrew names. HebrewCore adds nikud to Hebrew text, strips it again for matching, and transliterates in either direction between Hebrew and Latin script. Example prompt: Call POST `/v1/nikud` with a Hebrew sentence, then POST `/v1/transliterate` with the same text and direction he-to-latin ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/v1/translate` | Translate text | | POST | `/v1/translate/batch` | Translate a batch of strings | | POST | `/v1/detect` | Detect language | | POST | `/v1/nikud` | Add nikud | | POST | `/v1/transliterate` | Transliterate Hebrew and Latin script | | POST | `/v1/wrap` | Wrap a line for display | | POST | `/v1/medical/translate` | Verified clinical translation | | POST | `/v1/terms` | Upsert protected terms | ## Key resources - **Translation** — Single and batch translation with protection, glossary, gender, and domain options - **Language tools** — Direction analysis, language detection, transliteration, nikud, and display wrapping - **Account terms** — List, upsert, and delete protected terms kept in every protected translation - **Clinical API** — Verified clinical translation, SSE and WebSocket streaming, record-search queries, and verification ## AI readiness This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use. - **Score:** 74 / 100 - **Maturity:** AI-Aware - **Dimensions:** - Foundational Compliance: 99 / 100 - Developer Experience & Jentic Compatibility: 81 / 100 - AI-Readiness & Agent Experience: 53 / 100 - Agent Usability: 94 / 100 - Security: 60 / 100 - AI Discoverability: 98 / 100 - **View full report:** https://jentic.com/apis/itsbaba.com/itsbaba/scorecard - **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/ - **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview ### Score it yourself Every API in the directory is allowlisted, so you can re-score it with no key required. - **Score your own API:** https://jentic.com/scorecard.md - **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md ```sh npx @jentic/api-scorecard-cli score ``` ## Why Jentic - **Setup:** Wiring HebrewCore by hand means sending the bearer key, keeping hc_live_ and hc_test_ keys apart, and handling 429 rate_limited and quota_exceeded answers with their usage headers. Through Jentic you install once, import HebrewCore from the API Directory, store the key once, and your agent calls it. - **Permission scoping:** HebrewCore carries the text and options in the request body, so rules bound which operations your agent may call, such as translate and detect, not which text. You choose the operations it may call, so account changes like DELETE `/v1/terms/{id}` or the Clinical API are not included unless you add them. - **Credential handling:** Your HebrewCore API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'translate text into Hebrew' or 'add nikud', and Jentic returns the matching HebrewCore operation with its input schema, including the quality and protect options, so the agent calls the right endpoint. ## Related APIs - **Microsoft Translator** — General machine translation across many language pairs - **LibreTranslate** — Open-source, self-hostable machine translation - **Smartling** — Translation management for projects, files, and review workflows ## FAQ ### What authentication does the HebrewCore API use? Per its OpenAPI spec, every operation except GET `/v1/health` takes a bearer token: Authorization: Bearer with an hc_live_ key for production or an hc_test_ key for the sandbox. Through Jentic the key is stored by your own instance and attached at call time, so the agent never sees it. ### What are the rate limits for the HebrewCore API? The spec states 600 requests per minute per account across all its keys on every plan, answered with 429 rate_limited and Retry-After when exceeded. Monthly characters depend on your plan, and request size is capped at 20,000 characters, or 5,000 for transliterate and 10,000 for nikud. See https://hc.itsbaba.com/docs for details. ### Can HebrewCore keep numbers and product names unchanged in a translation? Yes. Set protect on POST `/v1/translate` and HebrewCore keeps prices, quantities, dates, codes, product names, and interface variables exactly as written. Pass protected_terms per request, or store them on the account with POST `/v1/terms` so every protected translation keeps them. ### Is there a HebrewCore MCP server? You don't need an MCP server to give your agent HebrewCore. Jentic connects it directly from the API Directory: import it, store your key once, and your agent calls the translation and language operations it needs without loading another server's tool definitions into its context. ### Can I limit what my agent is allowed to do with the HebrewCore API? Yes. Write a rule that allows only POST `/v1/translate` and POST `/v1/detect`, for example, and the agent can translate and detect language but cannot change or delete your account's protected terms or spend the higher-cost Clinical API characters. Every call it makes is logged. ### How do I translate text into Hebrew through Jentic? Search Jentic for 'translate text into Hebrew', load the POST `/v1/translate` operation, and execute it with text and target_lang he, adding quality or protect as needed. To run it on your own infrastructure, install Jentic One from its GitHub repo.