Chatdoc

What is ChatDOC?

ChatDOC is an AI-powered tool for reading, summarizing, and answering questions regarding PDF files and other types of documents. After uploading a document, users can pose questions, extract information, compare contents, and return to the relevant parts of the original text using citation markers.

The product supports not only individual PDF files but also documents in Word, Markdown, ePub, TXT formats, scanned images, pictures, and certain web page contents. It helps to improve reading efficiency, but it cannot replace manual verification of the original text, tables, formulas, and key facts.

Core functions

1. Document summary and Q&A

ChatDOC analyzes the structure of a document and generates summaries, suggested questions, and targeted answers. Users can ask it to summarize chapters, locate specific clauses, extract numbers, or explain technical concepts.

2. Traceable citations

The answers include citations to their sources; by clicking on or hovering over them, one can view the corresponding original text, tables, or images. Fine-grained tracking also allows users to jump directly to the relevant sections in the document from the key figures mentioned in the answers, thereby reducing the effort required for verification.

3. Multi-document and folder Q&A

Pro users can upload folders, organize multiple files into Collections, and pose questions across different documents. When exporting the Q&A pairs, the system also indicates which specific document the answer comes from.

4. Understanding of tables, formulas, and images

The product allows users to submit questions using text, tables, and formulas, and it can also interpret the content of images. However, complex multi-page tables, mathematical symbols, and low-quality scanned documents may still result in incorrect interpretations.

5. OCR scanning recognition

ChatDOC can process scanned PDF files and image-based documents; the latest update has made OCR available to all users. The number of pages that can be processed via OCR, as well as the supported languages and package limits, remain as specified on the account page.

6. Translation of the entire document

The product offers a full translation feature for PDF and DOC files, allowing documents to be converted into other languages. This feature is still marked as Beta; the translation of layouts, terminology, and tables requires manual review.

7. Custom models

Users can use their own model service keys to select compatible models from providers such as OpenAI, OpenRouter, DeepSeek, SiliconFlow, and Alibaba Cloud. Using one’s own keys entails costs associated with third-party models as well as data processing requirements.

List of main functions

• Summarize the entire document, chapters, papers, and reports.
• Ask questions about the document and receive answers with quotes from the original text.
• Select text, tables, formulas, or images for targeted analysis.
• Search, compare, and summarize information across multiple files.
• Identify scanned documents and extract the text content that can be used for answering questions.
• Identify contract terms, numbers, dates, parties involved, and key conclusions.
• Translate the entire PDF or DOC document into another language.
• Save frequently used prompts and reuse them in the reading workflow.
• Export Q&A from the document along with information about the corresponding source files.
• Access upload, Q&A, and PDF structuring capabilities through the API documentation.

Supported files and input methods

Which users are it suitable for

• Students and researchers who need to read papers, textbooks, and research reports quickly.
• Legal professionals who need to review contracts, case studies, and regulatory documents.
• Business and investment professionals who analyze financial reports, prospectuses, and industry reports.
• Engineers who search for product manuals, specifications, and technical documents.
• Consultants who need to compare and summarize multiple documents.
• Editing and content team responsible for organizing publications, facts, and background information.
• Developers who wish to integrate PDF parsing and document Q&A into their applications.

How to upload a document and start a Q&A session

1. Register an account and check the upload limits, number of pages, and file size constraints for the current plan.
2. Remove password protection from the document and confirm that you have the permission to upload it.
3. Click Upload or drag the file onto the page, and wait for the automatic parsing to complete.
4. First, check the system summary and document structure to confirm that the content is being read correctly.
5. Pose specific questions and request numbers, conditions, and sources.
6. Click on the citation to return to the original text and check whether the answer accurately captures the context.
7. Recheck the key conclusions using different phrasings, while retaining human judgment.

How to perform multi-document comparison

1. Place files related to the same project or topic in separate folders.
2. Create a Collection and upload folders or add items one by one.
3. Check the parsing status of each file and whether its name is correct.
4. First, ask about the core ideas of each document, and then pose questions for cross-document comparison.
5. The answer must indicate the source document, page number, and the quoted content.
6. For conflicting data, check the original file for the criteria and date.
7. After exporting the results, verify the references again; do not treat the AI-generated content as the original text.

Free version and Pro plan

The old help page for ChatDOC still indicates that the free version allows 2 files per day, with 20 pages per file; however, the more recent official updates state that the free version now permits 300 pages per file and 5 files per day, while the maximum file size is raised to 200MB. The documents available through different entry points have not yet been fully synchronized, so the information displayed on the account’s current page should be taken as the authoritative source.

Pro prices and additional packages

According to the official FAQ, the reference prices for automatic renewal are $8.99 for 30 days and $89.90 for 360 days; the older version of the blog showed higher original prices as well as promotional prices. Since payment methods and promotions can change, it is necessary to check the actual price and the terms of renewal on the upgrade page before making a purchase.

Payment methods include Visa, PayPal, and Alipay; some additional packages also support WeChat Pay. Pro renewals, issue packs, and file packs come with different quotas, which need to be confirmed separately at the time of purchase.

Model selection and built-in API key

• Built-in models can be used for summarization, suggesting questions, and document Q&A.
• The update log shows that GPT-5 mini is now an optional built-in model.
• DeepSeek-R1 supports the display of the inference process as well as the use of custom service providers.
• OpenAI and OpenRouter can be connected to a variety of compatible models.
• Providers such as SiliconFlow and Alibaba Cloud can be used to specify models.
• The cost associated with using custom keys is charged separately by the model provider.
• Different models have varying capabilities in terms of citation, image handling, context understanding, and structured output generation.
• The API Key should be stored securely in the account settings and must not be shared publicly.

ChatDOC API and PDF Parser API

ChatDOC offers developer APIs that enable file uploading, document-based Q&A, Collection management, and result retrieval. The PDF Parser API is designed to convert PDF files into structured JSON or Markdown format, and it also supports the extraction of document outlines and complex tables.

There are no fixed, universal figures for API prices and quotas; the values indicated in the developer console should be used as a reference. Developers also need to handle file permissions, task polling, retrying failed attempts, data deletion, and key security.

Why is manual verification still required for citations?

• Correct citation of a source does not mean that the AI has accurately interpreted the meaning of the original text.
• Tables may span multiple pages, have merged cells, or use different statistical criteria.
• OCR of scanned documents can mix up numbers, symbols, names, and technical terms.
• Negations, exceptions, and footnotes in the document are prone to being omitted in the summary.
• Multiple-file responses may combine content from different dates or versions.
• Legal, medical, financial, and academic citations must be verified against the original documents.

Product advantages

• The responses are combined with quotes from the original text, making the verification process clearer than that of ordinary messaging tools.
• It supports complex document elements such as text, tables, formulas, images, and scanned documents.
• The Collection and folder features are suitable for searching across documents and comparing them.
• It supports a variety of built-in models as well as service keys for user-defined models.
• Browser extensions can help with reading local or online PDF files in Chrome and Edge.
• It offers document Q&A, PDF parsing, and development interfaces such as Studio.
• The official GitHub provides API demos, OCRFlux, and development skills.

Usage restrictions and precautions

• There is a difference between the free quota specified on the official old help page and that in the latest logs.
• AI may produce incorrect summaries, erroneous interpretations of references, or overlook important terms.
• Complex tables, formulas, footnotes, and cross-page structures may still fail to be parsed.
• The fact that Pro does not impose limits on the number of PDF pages does not mean there are no restrictions regarding file size, tokens, or fair use.
• Advanced model packages, file packages, and membership quotas need to be purchased and managed separately.
• The entire translation is still in beta version; professional documents should not be delivered in this form.
• Custom models will send the content to the corresponding third-party provider for processing.
• The core code of the online platform is not open source, so it cannot be deployed privately in its entirety on one’s own.

Privacy and data security

ChatDOC processes uploaded documents via the cloud, and its privacy policy states that service providers are utilized to handle requests, account information is stored, and such processing takes place. According to this policy, personal information is generally not retained for more than 12 months after the account is closed, although legal, tax, or dispute-resolution requirements may necessitate a longer retention period.

• Do not upload confidential client information, unpublished contracts, or original documents containing personal data.
• When it is necessary to work with sensitive data, it is first necessary to anonymize that data and confirm that the organization permits the use of cloud services.
• Read the data retention and training policies of third-party model providers before using your own model keys.
• Before sharing a document or a Q&A link, check the access permissions and expiration date.
• Regularly delete files, Collections, and API keys that are no longer in use.
• Before installing a browser extension, check the permissions to read web pages and local files.
• High-risk industries should require written confirmation regarding compliance, data processing, and security.

Refunds and subscription management

• Refunds for eligible Pro purchases can be requested within 7 days.
• The official FAQ requires that the Pro benefits not be in use, along with the account details and payment proof.
• Refunds are typically processed within 3 to 5 business days after verification.
• Once automatic renewal is enabled, the fee for the next period will be charged in advance.
• If you do not wish to renew, you should cancel it on the Subscription page below the profile picture.
• Package files and issue packages have their own expiration dates, and these dates generally do not change when a member’s membership is suspended.

Official GitHub and open-source status

The official GitHub organization of ChatDOC has made available projects such as ChatDOC-API-Demo, API Reference, OCRFlux, and ChatDOC Studio Skills. OCRFlux is a tool for converting PDF files to Markdown format, while Demo and Skills are used for development purposes; these do not mean that the entire ChatDOC online platform is open source.

Platform and open-source status

Basic information

Recommendation score

4.5 / 5. ChatDOC’s functions for locating references, handling multiple documents in query-response scenarios, and processing complex PDFs make it suitable for professional reading, and its API set is quite comprehensive. However, document updates are not synchronized within the free tier, the additional packages are complex, and it is still necessary to consult the original text to verify important conclusions.

Frequently Asked Questions

Can ChatDOC be used for free?

Yes. The latest update log indicates that free users can upload 5 PDF files per day, with each file having a maximum of 300 pages, and files of up to 200MB in size are allowed; the information displayed in real time on the account should be taken as the final standard.

Does ChatDOC support scanning PDFs?

Supported. The latest logs indicate that OCR is now available to all users, but restrictions regarding the number of pages, languages, and quality still need to be checked on the account page.

Can multiple files be queried at the same time?

Yes. Pro supports uploading files to folders and handling multiple-document queries within Collections, and it can indicate the source file in the exported results.

Is ChatDOC a reliable source of answers?

Answers that include citations are easier to verify than those without any sources, but citations do not guarantee the accuracy of the explanations. Numbers, formulas, contracts, and professional conclusions still need to be checked manually.

Does ChatDOC provide an API?

Available. Developers can use the document Q&A API, PDF Parser API, and the interfaces related to ChatDOC Studio; the prices and usage limits are specified in the console.

Is ChatDOC open source?

The online platform is not fully open source; the official GitHub repository provides OCRFlux, API demos, documentation, and other supporting projects.