canonical: https://jentic.com/apis/auto-reader.com/auto-reader # Auto-Reader OCR + Translate API The Auto-Reader OCR + Translate API reads text from images and documents and translates it, with an Arabic-first and manga-aware recognition model. It detects and extracts text regions, runs OCR on images or base64 payloads, and can combine OCR and translation in a single call. Longer inputs run as asynchronous document and book jobs that a caller submits, polls, downloads, or deletes, and glossary, correction, and rewrite endpoints refine the extracted text. ## For AI agents Extract and translate text from images and documents: run multilingual OCR, detect and extract text regions, translate, and combine OCR with translation. Submit long documents and books as async jobs. Authenticate with a bearer key. ## Scope Does not generate images, synthesize speech, or store documents long term. Use for OCR, text extraction, and translation only. ## Capabilities - Run multilingual OCR on images, including Arabic-first and manga-aware recognition - Detect and extract text regions from an image before recognition - Translate supplied text, or run OCR and translation together in one call - Process a full document or submit a long book as an asynchronous job - Poll, download, or delete a submitted book or job - Set a translation glossary to keep terms consistent across a job - Correct, rewrite, and ask questions about extracted text ## Use cases ### Scanned Document Extraction An agent sends a scanned page to Auto-Reader, extracts the text regions, and runs OCR to turn the image into structured text it can act on. The Arabic-first model handles right-to-left scripts, so the agent processes multilingual scans without a separate recognizer per language. Example prompt: Call POST `/v1/detect` to find text regions, then POST `/v1/ocr` to extract the text from the image ### OCR and Translate in One Pass An agent takes a manga page or a foreign-language scan and returns readable translated text by calling the combined OCR-and-translate endpoint. Auto-Reader recognizes the source text and translates it in a single request, and a glossary keeps recurring names consistent, so the agent delivers a translated page without chaining two services. Example prompt: Call POST `/v1/ocr-translate` with the page image to extract and translate its text in one request ### AI Agent Integration via Jentic An AI agent reaches the Auto-Reader API through Jentic without holding the bearer key itself. It searches by intent, receives the matching operation with its input schema, and executes the call with the credential injected at run time, so OCR and translation become steps in a larger document workflow. Example prompt: Search Jentic for 'extract text from an image', load the POST `/v1/ocr` schema, and execute it with Jentic-managed credentials ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/v1/ocr` | Run OCR on an image | | POST | `/v1/manga` | Run manga-aware OCR on a page | | POST | `/v1/ocr-translate` | Run OCR and translation in one call | | POST | `/v1/translate` | Translate supplied text | | POST | `/v1/document` | Process a document | | POST | `/v1/book` | Submit a book job | | GET | `/v1/book/{job_id}` | Check a book job status | | PUT | `/v1/glossary` | Set a translation glossary | ## Key resources - **OCR and Detection** — Detect regions and extract text from images via `/v1/detect`, `/v1/extract`, and `/v1/ocr` - **Translation** — Translate text and combine OCR with translation via `/v1/translate` and `/v1/ocr-translate` - **Documents and Books** — Process documents and run book jobs via `/v1/document`, `/v1/book`, and `/v1/book/{job_id}` - **Refinement** — Keep terms consistent and refine output via `/v1/glossary`, `/v1/correct`, and `/v1/rewrite` ## 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:** 63 / 100 - **Maturity:** AI-Aware - **Dimensions:** - Foundational Compliance: 93 / 100 - Developer Experience & Jentic Compatibility: 62 / 100 - AI-Readiness & Agent Experience: 44 / 100 - Agent Usability: 94 / 100 - Security: 60 / 100 - AI Discoverability: 58 / 100 - **View full report:** https://jentic.com/apis/auto-reader.com/auto-reader/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 Auto-Reader by hand means managing its bearer key, choosing multipart or base64 upload for each image, and polling book and document jobs yourself. Through Jentic you install once, import Auto-Reader from the API Directory, store the key once, and your agent calls it. - **Permission scoping:** Auto-Reader carries the image or text inside each request body, so a rule bounds which operations your agent may call rather than which file it reads. You choose the operations it may use, such as OCR and translation, so the data-deletion operation is not included unless you add it. - **Credential handling:** Your Auto-Reader 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 'extract text from an image' or 'translate a document', and Jentic returns the matching Auto-Reader operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **Mindee** — Document AI and OCR for structured field extraction - **Cloudmersive OCR** — OCR for images and documents across many languages - **LibreTranslate** — Open machine translation between many languages ## FAQ ### What authentication does the Auto-Reader OCR + Translate API use? The Auto-Reader API authenticates with a bearer key in the Authorization header, per its OpenAPI spec. Through Jentic the key is stored encrypted by your own Jentic One instance and injected at execution time, so it never reaches the agent. ### Can I run OCR and translation in one call with the Auto-Reader API? Yes. POST `/v1/ocr-translate` extracts text from an image and translates it in a single request, while POST `/v1/ocr` and POST `/v1/translate` run each step on its own. A glossary set via PUT `/v1/glossary` keeps recurring terms and names consistent across a job. ### What are the rate limits for the Auto-Reader API? The Auto-Reader OpenAPI spec does not specify rate limits. Watch for HTTP 429 responses and check Auto-Reader's documentation for the current limits, and run long inputs as book or document jobs rather than many synchronous calls. ### How do I extract text from an image with the Auto-Reader API through Jentic? Search Jentic for 'extract text from an image', and it returns the Auto-Reader operation with its input schema. Load the schema and execute POST `/v1/ocr` with the credential injected at run time. To run it on your own infrastructure, install Jentic One from its GitHub repo. ### Is there an Auto-Reader MCP server? You don't need an MCP server to give your agent the Auto-Reader API. Jentic connects it directly from the API Directory: import it, store your key once, and your agent calls the OCR and translation operations it needs. Operations are discovered on demand, so no extra server's tool definitions sit in the agent's context. ### Can I limit what my agent is allowed to do with the Auto-Reader API? Yes. Auto-Reader carries the image or text inside each request, so a rule bounds which operations your agent may call. Allow only the OCR and translation operations the task needs, so the data-deletion operation stays off the table, and every call runs through your own instance where it is logged for review.