FileGPT

FileGPT is an AI-powered document and knowledge-base assistant designed to let users interact with information from multiple file and media formats through natural-language questions. Instead of manually searching through long documents, users can upload their files, select information sources, ask questions, and receive AI-generated answers based on the available content. The platform describes itself as an LLM-based application for interacting with content across different formats and focuses heavily on long-text processing and cross-source querying.

A central feature of FileGPT is multi-format content support. The official website lists PDFs, TXT files, DOC files, audio, YouTube content, and webpages among its supported sources. The platform is therefore designed to work with both traditional documents and online or multimedia information. This allows users to build a collection containing different types of content and interact with them from one conversation rather than switching between separate tools.

FileGPT’s Chat with Everything capability is designed around this multi-source workflow. Users can interact with files, audio, videos, YouTube content, and webpages through natural-language queries. Instead of treating every file as an isolated conversation, the platform lets users select multiple sources and ask questions across them, with the results combined into a single answer. This cross-source approach is one of the product’s defining features.

The platform also emphasizes long-text support. FileGPT specifically states that it supports large files and long texts, positioning this as a distinction from document assistants that impose relatively restrictive input sizes. On the current paid plans, individual file limits range from 100 MB on Basic to 200 MB on Pro, while the platform provides monthly token allocations for processing and querying content.

Another important feature is Smart Content Parsing. FileGPT says it can extract text from handwritten documents and analyze audio and video content. Handwritten-text support becomes available on Pro, while audio, video, image, and Bilibili analysis are also included on Pro. This extends the platform from ordinary text extraction into multimodal document analysis.

FileGPT also supports cross-source querying. Users can select multiple sources from a collection and ask a question that requires information from more than one file. This is useful when related information is spread across several documents, such as research materials, business files, project documentation, or collections of reference documents. The platform’s homepage specifically highlights the ability to integrate multiple files into one answer.

The product uses collections to organize information. Users can group files and other content sources together and then interact with the collection through chat. The current Basic plan supports up to 10 files per collection, while Pro raises this to 20 files per collection. Higher plans focus more on embedding the FileGPT experience into external websites and providing larger monthly token allowances.

FileGPT has a useful capability for organizations that want to expose their knowledge through a website. Website embedding is included from the Growth plan upward. The official site also provides an embedded FileGPT demonstration and states that users can embed their own bot containing videos, documents, or mixtures of content on their websites. This makes the platform relevant to organizations looking to provide an interactive knowledge interface around their own information.

The embedded approach can turn a collection of documents into a conversational knowledge experience. Rather than requiring a visitor to download a file and search it independently, a website owner can provide a FileGPT-powered interface in which visitors ask questions about the available information. The current Enterprise offering also removes FileGPT branding and adds enhanced analytics, providing a more organization-focused version of the same capability.

FileGPT also provides audio and video analysis. Pro users can ask questions about audio and video files, as well as images and Bilibili content. The service’s earlier Beta update announced full audio and video support alongside OCR and faster processing, showing that multimodal analysis has been part of the product since its early development.

Its support for handwritten documents is another specialized capability. FileGPT states that Pro users can extract and analyze handwritten text, which can be useful when working with scanned notes, handwritten records, forms, or other documents that are not available as ordinary machine-readable text. This is a more specialized use of document AI than conventional PDF chat.

The service is powered by large language model technology and provides different processing capabilities through a token-based system. FileGPT’s updates page explains that the service moved to a unified token model, while the current pricing page assigns specific monthly token allowances to each paid tier. These allowances increase significantly as users move from Basic through Enterprise.

FileGPT is designed for students, researchers, and business users. The Basic plan itself is described as being for students, researchers, and other users, while the higher tiers are positioned for individuals with larger needs, growth-focused use, and enterprise solutions. This gives the product a relatively broad audience spanning education, research, knowledge management, and document-heavy business workflows.

The service can also be useful for information retrieval and data review. Users do not necessarily need to read every page of a document before finding a relevant answer. Instead, they can ask a targeted question and use the resulting response to locate information in the source. This workflow is particularly practical when working with long reports, research documents, transcripts, or collections of business files.

FileGPT’s user reviews support this use case. G2 reviewers have praised the platform for quickly locating specific information across multiple file types, its clean interface, long-text support, and ability to preserve context and chat history. One reviewer specifically said the tool helped retrieve information from combined documents instead of requiring manual review of each file.

At the same time, the independent reviews identify some limitations. Several G2 reviewers reported slow performance, particularly on the free version, and some mentioned crashes or responsiveness issues. Other reviewers requested more model choices, better formatting, and more accessible support. One reviewer also noted that complex sections could occasionally be misinterpreted. These comments indicate that the user experience can depend on workload, document complexity, and the amount of processing required.

FileGPT has been in public development since 2023. Its terms and privacy policy both became effective on March 29, 2023, while the official updates page records Beta releases in April 2023. This makes 2023 a well-supported launch year for directory purposes.

Public company information is relatively limited. Third-party company directories identify Siyuan Peng as FileGPT’s founder/CEO and place the company in New York, while FileGPT’s own legal pages identify the service simply as FileGPT rather than publishing a detailed corporate profile. For a directory listing, FileGPT is the safest company-name entry rather than assigning an unverified legal entity name.

FileGPT’s current pricing uses a subscription and token-based model. The annual pricing page lists Basic at $100/year, Pro at $180/year, Growth at $480/year, Elite at $980/year, and Enterprise at $3,980/year. The website states that yearly plans include two months free. Basic includes 6,000 monthly tokens, Pro 12,000, Growth 36,000, Elite 75,000, and Enterprise 300,000 monthly tokens.

The plans are differentiated by file type, file size, collection limits, support, embedding, analytics, and processing capabilities. Basic supports full documents, webpages, and YouTube; Pro adds audio, video, image, Bilibili, unstructured PDFs, and handwritten-text support; Growth adds website embedding; Elite adds basic analytics; and Enterprise adds priority processing, enhanced analytics, and removal of FileGPT branding.

The website also states that users can try FileGPT for free, although the current public pricing page does not show a dedicated $0 plan with defined limits. For that reason, the free-access option is better treated as an entry-level free experience rather than assigning undocumented numerical limits to a formal Free plan.

FileGPT has an independent G2 rating of 3.8/5 from 9 reviews in the latest available G2 review page. Users most frequently praise ease of use, accuracy, the interface, long-text support, and cross-format information extraction. The main negative themes are slow performance, occasional crashes, limited model choices, and missing features.

FileGPT is therefore best suited to users who need to search, question, and analyze multiple files and other content sources from a single conversational interface. Its strongest differentiators are broad format support, cross-source querying, long-text handling, handwritten-text extraction, multimodal analysis, website embedding, and increasingly enterprise-oriented analytics and processing capabilities.

Pricing

• Basic: $100/year; 6,000 monthly tokens, 10 files per collection, 100 MB maximum single-file size.
• Pro: $180/year; 12,000 monthly tokens, 20 files per collection, 200 MB maximum single-file size, plus audio/video/image, Bilibili, unstructured PDF, and handwritten-text support.
• Growth: $480/year; 36,000 monthly tokens and website embedding.
• Elite: $980/year; 75,000 monthly tokens, website embedding, and basic analytics.
• Enterprise: $3,980/year; 300,000 monthly tokens, website embedding, priority processing, enhanced analytics, and removal of FileGPT branding.

The official pricing page states that annual plans include two months free.

Review

FileGPT currently has a 3.8/5 rating from 9 G2 reviews. Reviewers commonly praise the clean interface, ease of use, accuracy of information retrieval, long-text support, and ability to work across multiple file formats. Several reviewers have also reported slow performance, occasional crashes, limited model choices, and areas where formatting or support could be improved.