Google AI Releases LangExtract: An Open Source Python Library that Extracts Structured Data from Unstructured Text Documents

In today’s data-based world, valuable ideas are often buried in an unorganized text-whether this is clinical notes, long legal contracts or clinical notes indicators. It is a technical and practical challenge that can be tracked from these documents. Langixtract, Lankextract, the new Google AI’s Python Library, is designed to process this gap directly, using LLMS like Gemini to provide strong automated automatic extraction with trace and transparency in its essence.

1. Extracting the permit and can be followed

Langextract allows users to determine dedicated extract tasks using natural language instructions and “low -shot” examples. This enables developers and analysts Select exactly the entities, relationships or facts that must be extracted, and in any structure. Decally, each part of the extracted information It was linked directly to the text of the sourceCheck the health and review of the tracking.

2. The ingenuity of the field

The library works not only in technology offers but in critical areas-including health (clinical notes, medical reports), financing (summaries, risk documents), law (contracts), research literature, and even arts (Shakespeare analysis). The original use cases include automatic extraction of medicines and doses and management details from clinical documents, as well as relationships and emotions from plays or literature.

3. Planning outlet with llms

Silver is compatible with other LLMS, enabling Lankxtract The enforcement of allocated output plans (Like Json), so the results are not only accurate – they are immediately used in databases, analyzes, or artificial intelligence tubes. It solves the traditional LLM points around hallucinations and the planned drifting by grounding outputs to both user instructions and the actual source text.

4. The ability to expand and perceive

• He deals with large folders: It treats LankExtract efficiently long documents by minimizing results, parallel and assembly.
• Interactive perception: Developers can create an interactive HTML reports, display each extracted entity with a context by highlighting its location in the original document – checking and analyzing smooth errors.
• Smooth integration: It works in Google Colab, Jupyter, or as independent HTML files, and support a quick reaction loop for developers and researchers.

5. Installation and use

Easily installation with PIP:

Example workflow (extracting letter information from Shakespeare):

import langextract as lx
import textwrap

# 1. Define your prompt
prompt = textwrap.dedent("""
Extract characters, emotions, and relationships in order of appearance.
Use exact text for extractions. Do not paraphrase or overlap entities.
Provide meaningful attributes for each entity to add context.
""")

# 2. Give a high-quality example
examples = [
    lx.data.ExampleData(
        text="ROMEO. But soft! What light through yonder window breaks? It is the east, and Juliet is the sun.",
        extractions=[
            lx.data.Extraction(extraction_class="character", extraction_text="ROMEO", attributes={"emotional_state": "wonder"}),
            lx.data.Extraction(extraction_class="emotion", extraction_text="But soft!", attributes={"feeling": "gentle awe"}),
            lx.data.Extraction(extraction_class="relationship", extraction_text="Juliet is the sun", attributes={"type": "metaphor"}),
        ],
    )
]

# 3. Extract from new text
input_text = "Lady Juliet gazed longingly at the stars, her heart aching for Romeo"

result = lx.extract(
    text_or_documents=input_text,
    prompt_description=prompt,
    examples=examples,
    model_id="gemini-2.5-pro"
)

# 4. Save and visualize results
lx.io.save_annotated_documents([result], output_name="extraction_results.jsonl")
html_content = lx.visualize("extraction_results.jsonl")
with open("visualization.html", "w") as f:
    f.write(html_content)

This produces JSON’s organized outputs and source source, as well as an interactive HTML perception for easy review and clarification.

Realistic specialized applications

• drug: Eliminate medicines, doses and timing and link them to return to the source of the source. Supported by visions of research conducted on accelerating medical information extraction, the Langextract approach applies directly to the structuring of clinical and radiology reports – defining clarity and supporting interconnection.
• Finance and lawAutomatically withdraws sentences, terminology, or risks from the dense legal or financial text, ensuring the possibility of restoring all the output to its context.
• Search and extract dataCoordination of high lines extract from thousands of scientific papers.

Even the team offers a demonstration called Radextract The structure of the radiology reports – not only highlights what was extracted, but exactly as information appeared in the original inputs.

How to compare Lankixtract

In summary

Lankxtract introduces a new era to extract organized and enforceable data from text – concrete:

• The permit, extraction can be explained
• Results that can be tracked with the support of the source context
• Immediate perception of rapid repetition
• Easy

verify Jaytap page and Technical Blog. Do not hesitate to check our GitHub page for lessons, symbols and notebooks. Also, do not hesitate to follow us twitter And do not forget to join 100K+ ML Subreddit And subscribe to Our newsletter.