Changelog

🚀Features

• Added a Notifications tab to the dashboard. You can now see when your processing jobs are finished, making it easier to track job progress without constantly refreshing. Email notifications for job completion are coming soon.

⚡Improvements

• Touchup to the Review UI by removing visual clutter. The confidence counts bar and highlight toggle have been removed for a cleaner experience - highlights are now always enabled by default.

• Added a Careers page. If you like building awesome things, and understand that AI is going to transform everything and must be embedded into every action you take - give us a shout!

🚀Features

• Added search functionality to the Documents and Standardizations dashboards. You can now search by filename, document ID, or standardization ID using case-insensitive partial matching.

• Launched a new interactive schema creation experience. The new flow features a chat-based assistant that helps you design your schema by understanding your extraction needs. You can preview the schema and see sample extraction results in real-time, iterate on the design through conversation, and swap sample documents to test different scenarios.

⚡Improvements

• Web UI: added the ability to mass-download original PDFs, as well as PDFs with an OCR layer (that means you can control+F search the documents). Simply multiple-select files, hit download, and you're off to the races!

⚡Improvements

• Webhook payloads that exceed size limits are now handled gracefully. When document.processed.success or standardization.processed.success webhooks are too large, they will be retried with the large field (text or data) omitted, and a new flag (textTooBig or dataTooBig) set to true. Use GET /document/{documentId} or GET /standardization/{standardizationId} to fetch the full content when this occurs.

⚡Improvements

• Upgraded our infra-scaling for increased reliability. Previously, when under heavy load, DocuPipe was more prone to dropping jobs while the system spun up new machines. With this update, scaling is done more carefully to avoid drops that lead to 'processing timeout' errors.

• Downloaded OCR-layered PDFs now have all pages automatically rotated into upright orientation for easier reading. Any page that is upside-down or sideways in the original document will be rotated to upright in the searchable PDF version.

⚡Improvements

• Added support for multi-image TIFF uploads.

💰Pricing

• We have changed again the way we calculate the number of pages for tabular documents (CSV, Excel), resulting in an additional 50% decrease in page count for such documents. The new pricing is effective immediately for all newly uploaded tabular documents. You will see the calculated number of pages to be around 50% of what it was, and 30% of the original calculation before November 4, 2025.

🔧Fixes

• Developers: all DocuPipe users now have access to new webhook event types: review.verified.success fires when a review is verified (approved) by a human reviewer; review.rejected.success fires when a review is rejected by a human reviewer. This lets you explicitly trigger action on final human review of objects, where needed

💰Pricing

• We have changed the way we calculate the number of pages for tabular documents (CSV, Excel), resulting in a roughly 40% cost decrease for standardizing such documents. The new pricing is effective immediately for all newly uploaded tabular documents. You will see the calculated number of pages to be around 60% of what it was before.

🚀Features

• Extended 🧠 standardization level is now available as an experimental mode. It’s more compute-intensive and slower to run, costing 5 credits per page - a number that is subject to change in the coming weeks. It handles ambiguous or complex cases better. Improvement is especially felt in long, cross-page context such as multi-page tables or extractions that need to look at multiple pages to make sense of a single item

• To use extended effort, simply set effort level to "extended" from the drop down when you click on standardize, or change the `effortLevel` parameter in our API.

🚀Features

• Added a new Parse → Split → Classify → Standardize workflow step that is available through both the API and the website dashboard. Workflows can now split documents into multiple files based on content, then classify each resulting file, and finally standardize them with different schemas based on the classification results.

🚀Features

• Added a new Parse → Split → Standardize workflow step that is available through both the API and the website dashboard. Workflows can now split documents into multiple files based on content, and then standardize each resulting file with specified a schema

• Added a schema comparison viewer to the Schemas dashboard. Select exactly two schemas to open a side-by-side diff that highlights JSON schema and guideline changes.

💰Pricing

• Generating schemas with the Improve chat assistant now costs 1 credit per page, matching the pricing of the Create Schema flow. The chat itself is still free to use, credits are only charged when you initiate schema generation.

🚀Features

• Overhauled the Improve Schema experience with a chat-based assistant that helps you diagnose standardization issues, proposes targeted schema updates, and generate fresh schemas on demand. The new flow is completely free (no credit usage) while it remains in beta, and we'll continue polishing it over the coming weeks.

• Bulk Excel download functionality has been added to the Standardizations dashboard. Users can now select multiple standardizations and download them as Excel files in a single ZIP archive. Previously, Excel downloads were limited to single standardizations only.

• Added 'Display Mode' option for document classification, available in both the API and website dashboard. You can now choose between auto, spatial, sections, or image display modes when classifying documents. This allows you to control how the document is presented to the AI, improving classification accuracy for specific document types. The image mode is particularly powerful - it enables classification of any image (photos, diagrams, etc.) even if it's not a traditional document.

🚀Features

• Added a new Parse → Standardize → Review workflow step that is available through both the API and the website dashboard. Workflows can now standardize documents and immediately create a review object, giving teams an end-to-end automation option for validating results.

🚀Features

• Workflows are now available on the website dashboard (previously API-only). Workflows allow you to automatically trigger standardization or classification actions immediately after document parsing is complete, creating streamlined processing pipelines. You can create workflows that either standardize documents with specific schemas, or first classify documents and then apply different schemas based on the classification results. The new Workflows dashboard lets you create, edit, view, and directly apply workflows to document uploads.

🚀Features

• Added overage control setting in Account Settings. Paid tier users can now disable overage charges to prevent unexpected billing. When disabled, API requests that would exceed plan limits will be rejected instead of incurring overage charges. The setting is enabled by default and can be toggled in the General settings tab.

🚀Features

• Added Merge functionality to combine multiple documents into a single document. Users can select documents from the dashboard and specify the order of merging.

⚡Improvements

• Document parsing now supports XFA-based PDFs, which are commonly used for interactive forms. Previously, these documents would fail to parse. This improvement allows users to upload and parse a wider range of PDF documents, including those with dynamic content.

• Added a `timeout` arg to document upload and standardization, which allows the failure webhook to fire if a result is not ready within the specified time (in seconds). This is useful for API and Make users who want to specify a maximum wait time for results, and have the webhook fire if the time is exceeded so they can retry or handle the timeout appropriately.

• Analyze now has an additional input parameter `instructions`, which allows you to provide custom instructions to the AI on how to analyze the document or dataset, giving you more control over the output and establishing a global context for the analysis.

• Analyze via the website now has a quality of life feature that lets you prepopulate the questions + instructions inputs from previously completed jobs, saving time of manually retyping them.

💰Pricing

• Merge priced at 0.01 credits per page (rounded up).

🚀Features

• Added Query functionality to search standardized documents using natural language, at a flat fee of 2 credits per query.

• Use plain language to describe what you are looking for, and it will search across all standardized documents to find the relevant matches.

⚡Improvements

• Downloading searchable PDFs now works in Hebrew, following popular demand. This allows users to download PDFs with the OCR layer baked in, which can be searched using CTRL+F or similar functionality in PDF readers, or copy/pasted.

🚀Features

• Added Display Mode selection to the Split functionality. Users can now choose how documents should be displayed during processing from four options: Auto, Spatial, Sections, or Image.

🚀Features

• Added API key revocation and reissue functionality in the Settings page. Users can now revoke their current API key and generate a new one instantly from the Danger Zone section. This improves security by allowing users to invalidate compromised keys immediately.

🚀Features

• Bulk XML download functionality has been added to the Standardizations dashboard. Users can now select multiple standardizations and download them as XML files in a single ZIP archive. Previously, XML downloads were limited to single standardizations only.

🚀Features

• A new parse version (V3) has been released, which makes minor improvements in how `spatial` representation is performed. This is now the default on the website and API.

• Standardization V2.2 has been set as the new default in the API.

🚀Features

• AI Review now localizes individual lines and words in the document as evidence. Previously we only identified entire sections. You can now speed through reviews by focusing on the most relevant parts of the document.

• Confidence estimates in AI reviews are now significantly more accurate, allowing you to build workflows that focus on low-confidence items.

• AI Review now supports a human-in-the-loop editing. You can now add, change, and remove values in the review interface, and set an official "verified" status for items. This supports flows that require human validation of AI results, such as in legal or compliance contexts.

📣News

• We're still focusing our attention on making review workflows more useful. Please reach out to customer support if you run into any issues or have a feature request to make review capabilities more useful to you.

🚀Features

• AI Review has been completely overhauled! AI review lets you see where in a document results came from.

• Features include: handling of very long documents and thousands of standardization items. You also get much more precise localization, and vastly improved reliability

• Review now lets all users change, remove, and add values, as well as set an official "finalized" state.

• This allows you to seamlessly build human-backed pipelines that can be used to validate results, improving your document reviewing humans 10x faster than before.

📣News

• Standardization V2.2 will become the new default on the API in one week, on June 12, 2025. This version has been available for a while now, and we have received positive feedback from users. It is faster and more accurate than the previous version, and we encourage everyone to try it out. There might be several final tweaks before stabilization as the new default. To use a previous version, simply set stdVersion=2.0 or 2.1 in the API.

🚀Features

• Standardization V2.2 has been released on both the website and API. No interface or parameter changes, but the model has been improved to better handle complex documents and edge cases, and should generally produce more accurate results. The default version on the API is still V2.1, but you can select the new version explicitly by setting stdVersion=2.2. The website is now using V2.2 by default.

• Expanded file‑format support – Tables (.csv, .xls, .xlsx) uploads are now supported via both the website and API. Table files are parsed natively and do not consume parsing credits.

⚡Improvements

• DocuPipe now supports larger document uploads. The new maximum file size is 2GB (up from 1GB), and the maximum number of pages is 1000 (up from 500). This change is available via both the website and the API. We still recommend exercising caution when standardizing large documents, as job failures are more likely.

• Analysis results are now viewable from the Jobs tab instead of needing to navigate to the Analysis tab.

• Document viewer now has zoom control, to allow for easier reading of documents. Quality of life!

📣News

• We’ve officially rebranded from DocuPanda to DocuPipe. There’s no impact on service – all existing functionality remains the same. The only change is that our website and API endpoints now use docupipe.ai. The previous API URL (app.docupanda.io) will continue to work without changes. We’re excited to continue evolving under our new name and bring you even more powerful features ahead.

• We’ve entered the 3-month observation period for our SOC 2 Type II audit and expect full certification by August 2025. Security and privacy have always been core to DocuPipe, and this milestone underscores our ongoing commitment to maintaining the highest standards.

• Accordingly, our Trust Center has been updated to reflect this change. You can find the new Trust Center at https://trust.docupipe.ai. This page provides very comprehensive information about our security and privacy practices and tech stack.

🚀Features

• When uploading a PDF, you can now specify the page range to be parsed. This is useful if you only want to upload a specific subset of pages, and only get charged credits for those pages.

• It is now possible to standardize without a schema from the website (previously it was only possible via the API). This is useful if you want to standardize a document without having to create a schema first. The schema will be created automatically based on the document content and the custom instructions (guidelines).

• Split now has two additional quality-of-life input parameters: dataset and filename prefix. Both of these new arguments are optional, and leaving them blank keeps the previous behavior. The dataset parameter allows you to assign a new dataset value to any documents generated from the Split job, and the filename prefix parameter allows you to control what prefix will be given to newly generated files (the suffix is still determined by the AI, as before).

🚀Features

• Standardize V2.1 released.

• The only difference is that fields which weren't found are now an explicit null. For example, we will produce {"rentAmount": null} where before we used to omit rentAmount alltogether if not found.

• Scheduled Change: Version 2.1 will become default for all API users and website visitors. This will happen April 23, 2025.

• If you must retain the old behavior, you can set stdVersion=2.0 in the API.

• Expanded file‑format support – Word (.doc, .docx), HTML, plain text, and JSON uploads are now supported via both the website and API. Word and HTML files are converted to PDF; plain text and JSON files are parsed natively and do not consume parsing credits. Excel support is coming soon.

⚡Improvements

• Parallel batch processing: Standardize batch jobs now execute documents in parallel instead of sequentially, significantly reducing overall processing time.

💰Pricing

• The promotion period for high effort level standardization ends April 23, 2025. From that date, setting effortLevel='high' will be charged 4 credits per page. This is double the standard rate of 2 credits per standardization.

• The promotion period for review action ends April 23, 2024. From that date onwards, the review action will be charged 2 credit per page. This was previously free for our beta users.

🚀Features

• Standardization V2 is now the new default in the website. We will continue supporting all versions on the API and website, but encourage users to try the new version. It will continue improving in the coming weeks, and we appreciate the feedback we have been getting from users to help improve it and make it even better.

⚡Improvements

• The document viewer in the website has received a makeover, allowing you to more seamlessly move between viewing your document, text results, and standardizations. We hope you enjoy!

• Uploading documents now auto-suggests relevant schemas based on the document type, allowing for a more streamlined flow.

💰Pricing

• The cost of schema creation and refinement has been reduced from 2 credits per page to 1 credit per page. We think this will let users feel more free to create more schemas and iterate on them in the Improve tab.

• The cost of classification has been reduced from 1 credit per document, to 0.1 credit per page (rounded up), as this is more fair for most users.

• The cost of analyze has been reduced to 0.5 credit per page (rounded up), and has been simplified into a uniform pricing regardless of whether you are analyzing a single document or multiple documents. The behavior has also changed to be more verbose.

🚀Features

• We are pleased to announce that Standardization V2 has been released! It is available via both the website and the API. The new version is faster, more accurate, and more flexible than the previous one. However, it might not immediately work better for everyone, and V1 is still available and will remain the default on the website for a while longer. From the API, Standardization V2 has a separate endpoint: /v2/standardize/batch (see API docs for details). Also see a new article on Standardization in the Help Center that explains a bit about how the new version works, and what its input parameters are for. Please start experimenting with V2 and give us feedback, it will still be improving over the next weeks before the launch is finalized and it becomes the new default.

⚡Improvements

• The parsing model has been upgraded to V2 in the API (in addition to the website). The POST /document endpoint accepts an optional parseVersion parameter, which can be set to 1 or 2 (default is now 2).

• We have improved how we spatially display documents to the AI in standardization and analysis, which should improve results.

🚀Features

• Standardizations can now be downloaded as Excel files from the API as well, under the endpoint /standardization/{standardization_id}/download/excel-url, which gives you a temporary URL to download the Excel file. This feature is free of charge.

⚡Improvements

• The parsing model has been upgraded to a new version (V2), which improves accuracy with tables, checkmarks, and handwriting recognition. This update is now the default on the website and will become the default in the API in one week. The POST /document endpoint now accepts an optional parseVersion parameter, which can be set to 1 or 2. The default remains 1 for now but will switch to 2 in one week. To continue using the old version, set parseVersion to 1.

🚀Features

• We added an ability to download individual standardizations as an Excel file. This feature is currently available only via the website, under the Standardization tab: click Download -> Excel. The Excel file will contain the same information as the standardization details page, but in a more structured format: non-array fields will be in a sheet called 'main', and array fields will be in separate sheets named after the array field. This feature is free of charge, and will be available in the API soon.

🔧Fixes

• In document parsing, we removed the underscore padding in tables, as it caused issues with some documents. Newly parsed documents will revert to the previous behavior of having empty table cells filled with a simple empty string. For standardization with standardizationMode='sectionBased', we will still use padding to improve results.

⚡Improvements

• You can now right-click tabs in the dashboard menu for opening a new tab (this was not possible before).

• We have disabled the mobile view for the dashboard, as it was not optimized for mobile devices and caused issues.

🚀Features

• We added an API endpoint to POST a new schema from scratch. Up until now, a schema could only be updated from an existing schema, but now you can add a schema object directly using the API. The endpoint is POST /schema - find more details in the API docs.

🔧Fixes

• A bug was fixed where previously we allowed schemas to have fields with type=enum, which is not a valid type in JSON Schemas (enum is an additional key in a field, not a type). We only allow the types 'string', 'number', 'integer', 'boolean', 'object', 'array'.

🚀Features

• Added the ability to download a PDF with the OCR layer baked in. Available both in the API at the endpoint document/{document_id}/download/ocr-url or via the website, under the Documents tab: click Download -> File (OCR Layer). In further detail, this feature allows you to download your PDF - which may be handwritten or contain images - with DocuPipe's OCR layer placed on top of the document in invisible font on the word level. This allows you to search your PDF, or highlight / copy text from it, even if the original document was just a scan. This service is free of charge.

⚡Improvements

• In document parsing, we added underscore padding in tables (instead of empty string), which improves readability / rendering and standardization results, as it makes it easier for the AI to keep track of table columns. This affects anyone using the document.result.text output in its raw form, or anyone using standardization with standardizationMode='sectionBased'.