Unstract(AGPL) – Launch LLM-powered APIs to structure unstructured docs
Hello HN! One of the most common uses of LLMs is to go beyond what traditional RPA or IDP can do when it comes to structuring unstructured documents. However, there are a lot of challenges in getting this done right from extraction of text data from PDFs, scanned images or other formats, prompt engineering, evaluation and integration with existing systems. This very specific use case is where Unstract can help teams move really fast, leveraging LLMs. By doing the heavy-lifting in this fast-changing ecosystem it lets engineers concentrate on implementing core business workflow automations.…
In plain words
Unstract is an open-source platform for extracting and structuring data from unstructured documents like PDFs and scanned images using large language models. It handles text extraction, prompt engineering, and system integration, allowing developers to build LLM-powered APIs without managing low-level technical details. The tool is released under AGPL and can be deployed quickly using Docker Compose, enabling teams to automate document processing workflows faster than traditional RPA or intelligent document processing solutions.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hello HN! One of the most common uses of LLMs is to go beyond what traditional RPA or IDP can do when it comes to structuring unstructured documents. However, there are a lot of challenges in getting this done right from extraction of text data from PDFs, scanned images or other formats, prompt engineering, evaluation and integration with existing systems. This very specific use case is where Unstract can help teams move really fast, leveraging LLMs. By doing the heavy-lifting in this fast-changing ecosystem it lets engineers concentrate on implementing core business workflow automations. Here's what's special about Unstract: AGPL-licensed: Unstract is available under the AGPL, which is friendly for personal or commercial use. Getting started is simply a matter of running a script, which should take care of everything leveraging Docker Compose. We've also made is very easy to update the container images should you wish to. If you have a Mac or a Linux laptop, you can get started in minutes. LLMWhisperer: Results from LLMs are as good as the input they receive. LLMWhisperer is a cloud service that given PDFs (native text, scanned) returns text data in a way LLMs are best able to "understand". LLMWhisperer is free for up to 100 pages a day. While Unstract supports LLMWhisperer, we're also adding support for unstructured.io and LlamaParse as well (they're both great services!). For more details: https://llmwhisperer.unstract.com Strong, improving ecosystem support: Unstract supports a variety of providers for LLMs, Vector Databases, Embeddings, Cloud File Storage systems and databases/data warehouses. A full list is available on our Github page (https://github.com/Zipstack/unstract). No-code Prompt Studio: You need a way to figure out how well your prompts are working for a set of documents to make sure the prompts you're writing are as generic as possible. This is not easy to do with a regular IDE and a file manager on your laptop. Prompt Studio a purpose-built, no-code environment that makes it easy for you to develop, run and debug prompts while referring to sample documents side-by-side. The whole project is stored as a YAML (export/import options coming soon). Easy integrations: Unstract supports the launching of APIs and ETL Pipelines that take in unstructured documents and produce JSON data or sync data with various data sources like Snowflake, Bigquery, Redshift, PostgreSQL, etc. Prompt Studio no-code projects can be launched as APIs or ETL Pipelines with ease in just minutes letting you leverage LLMs to solve business automation problems. Unstructured data APIs are useful when an existing application needs to automate the processing of unstructured docs. Unstructured data ETL Pipelines are useful when unstructured docs in cloud file or object stores need to be structured and pushed into databases or data warehouses. Quick start: See our documentation (https://docs.unstract.com) to walk through a quick start project structuring standard credit card statements from a couple of different banks. You can also view a 3-min video of Unstract here: https://www.youtube.com/watch?v=8aZh-pRwZh8 Enterprise-only features: Unstract has a couple of enterprise/cloud/managed features that are not part of the open source offering: LLMEval uses more than one LLM to arrive at a consensus on extracted fields. If there is no consensus, the field is set to null (a null value is always better than a wrong value), after which a human can review the extraction results. There are a couple of advanced Prompt Studio features that can cut token usage (and thus costs) by up to 6x (Single Pass and Summary-based Extraction). Otherwise, all other features are available in our open source version.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, March 2024
the whole month →
- 3Y
Life & fun · 2024 · github.com
Microlaunch▲1,116Launch and get feedback on both the idea and product
Dev tools · 2024 · microlaunch.net
