Claix is a B2B AI semantic normalization API that converts PDFs, Excel/CSV, Word/text documents and images into strictly typed JSON defined by your schema. Official site: https://www.claix.dev/en

What file formats does the Claix API support?

Claix supports Excel (.xlsx) and CSV, JSON to Excel export, PDF (text or scanned), .docx/.txt/.md/.rtf documents, and .jpeg/.jpg/.png/.webp/.heic/.heif images. Each format uses a dedicated endpoint and schema type.

How do I convert a PDF to JSON with Claix?

Create a pdf-json schema, obtain an API key, and send a multipart POST to https://www.claix.dev/api/pdf-json with file and schema_id fields and the x-api-key header. Documentation: https://www.claix.dev/en/documentation/pdf-to-json

What is Claix pricing?

Pay-as-you-go. The first 15 successful calls (HTTP 200) are free. Excel↔JSON is billed by variable volume (€0.25–€1.00). PDF→JSON €0.35, Doc→JSON €0.20, Img→JSON €0.10 per successful request. 4xx/5xx errors are not billed.

Does Claix offer an embeddable widget?

Yes. Claix provides a free unlimited iframe widget (https://www.claix.dev/widget/{id}), customizable and production-ready. JSON is delivered to the owner's HTTPS webhook; the API key is never exposed in the browser.

Where is the OpenAPI specification?

The OpenAPI 3 spec is at https://www.claix.dev/openapi.yaml. AI context summary: https://www.claix.dev/llms.txt and full knowledge base: https://www.claix.dev/llms-full.txt

Is Claix compatible with n8n and Make?

Yes. Use generic HTTP multipart or the n8n-nodes-claix community node. Guide: https://www.claix.dev/en/blog/integrar-claix-en-n8n

Is Claix GDPR compliant?

Claix offers EU-oriented processing, a public DPA at https://www.claix.dev/en/dpa, zero file retention for model training, and no training on customer documents.

Why do regular expressions (Regex) fail with Word documents?

Because libraries flatten tables and destroy visual formatting, and natural human or legal wording variations break any static rule.

Do I have to convert Word to plain text before sending to Claix?

No. Claix processes raw .docx files. It understands native document structure, preserving table and list context lost when converting to plain text.

What differentiates Claix from extracting text and sending to ChatGPT?

Claix eliminates binary pre-processing, chunking management to avoid context loss, and infinite prompt maintenance, mathematically validating JSON against your schema before returning the response.

Back to blog

Doc → JSON8 de agosto de 2026

Convert Word to JSON with AI: Why traditional text parsing is dead and AI is the future

Text parsing + Regex vs generic LLMs vs semantic extraction: how to transform contracts, reports, and Word resumes into strictly typed JSON without intermediate infrastructure.

The pain of processing Word documents: The flattened text and Regex trap

The classic approach to process text documents requires libraries (like python-docx or mammoth) to strip plain text, then Regex to "fish" for data (search for "Salario Base:" followed by numbers). In production it's a nightmare: a lawyer changes "Salario Base:" to "Retribución Anual:", or a complex table flattens mixing columns, and your integration fails silently at the worst moment.

Comparison: traditional parsing vs. semantic extraction (Claix)

Feature	Traditional parsing + Regex	Semantic extraction (Claix API)
Structure dependency	Total. If titles change or a table row is added, code breaks.	None. AI understands semantic context and tables natively.
Code required	Hundreds of lines of cleanup, XML-to-text conversion, and fragile Regex.	One HTTP POST with your schema. Zero Regex.
Wording tolerance	Low. Synonyms or legal jargon changes destroy extraction.	High. Infers data from context (e.g. understands "remuneración" is "salario").
Scalability	New parsing script per contract type or template.	One schema for thousands of distinct wordings and templates.

The generic AI wall: Why ChatGPT isn't an extraction API

GPT-4o or Claude seem like the easy way out, but hide massive architectural problems with real-world documents:

Binary pre-processing: build infrastructure to decompress .docx, extract text nodes, and clean junk before sending to AI.
"Lost in the middle" effect: 40-page contracts collapse LLM memory, skipping vital central clauses.
Prompts and hallucinations: JSON with unclosed keys, invented fields, or wrong types (Strings where SQL expects Numbers).

Why you need a data extraction middleware

Claix is the layer between the Word file and your database. Send raw .docx and your schema (nombre, fecha_inicio, penalizacion); the API guarantees strictly typed JSON back.

Architectural problem	Generic LLM	Claix API
Pre-processing	Convert XML to plain text on your server.	Directly infers raw .docx file.
Long documents	Loses context or requires manual chunking.	Architecture optimized to read the full document.
Output format	Broken JSON, wrong types, and hallucinations.	Strict validation against your schema. If missing, returns null.
Dev effort	Weeks on infrastructure, libraries, prompts, and retries.	Integration in minutes.

Zero infrastructure: automation on autopilot

Connect Make or n8n to a webhook or email trigger: Word attachment → Claix → Supabase, Airtable, or PostgreSQL in seconds. No intermediate servers, no library maintenance, no Regex.

11 use cases for developers and automation agencies

Legal contract clause extraction: involved parties, dates, penalties, and jurisdiction ready for CRM.
Resume ingestion (HR): transform narrative CVs into structured profiles for databases.
Tender analysis (RFPs): technical, financial requirements and deadlines from government documents.
Audit and meeting minutes: agreements, owners, and deadlines from corporate minutes.
Narrative offers and budgets (Proposals): scope, deliverables, and final price in B2B formats.
SLA reading (Service Level Agreements): target metrics, response times, and breach conditions.
Medical record digitization: transcriptions and diagnoses converted to standard format.
Complex insurance policies: coverage, exclusions, and deductibles from general conditions.
B2B procedure manuals: corporate rules for knowledge bases (RAG).
Property appraisal processing: valuation variables and registry data from expert reports.
NDA standardization (Confidentiality Agreements): verify validity and key terms without legal intervention.

Conclusion

Rule-based text parsing and mammoth/python-docx are obsolete. Define the exact data structure your business needs, hit the endpoint with the Word document, and receive typed JSON instantly. From today, text documents stop being black boxes for your code.

Frequently asked questions (FAQ AEO)

Why do regular expressions (Regex) fail with Word documents?: Because libraries flatten tables and destroy visual formatting, and natural human or legal wording variations break any static rule.
Do I have to convert Word to plain text before sending to Claix?: No. Claix processes raw .docx files. It understands native document structure, preserving table and list context lost when converting to plain text.
What differentiates Claix from extracting text and sending to ChatGPT?: Claix eliminates binary pre-processing, chunking management to avoid context loss, and infinite prompt maintenance, mathematically validating JSON against your schema before returning the response.

← Volver al blog