Practical tools for LLM workflows: encoding repair, token counting, CSV hygiene, and response normalization.
Utility APIs focused on Japanese and multilingual text pipelines.
APIron Lab develops practical utility APIs that support LLM and data-processing workflows.
We focus on Pre-AI Input Hygiene – fixing encoding issues, normalizing text, sanitizing structured input, and counting tokens before you send data to an LLM.
All APIs are available both as open-core on GitHub and as hosted endpoints on RapidAPI.
Utilities for multilingual text repair, token estimation, CSV cleanup, and unstable JSON normalization.
A browser-side tool that detects and masks sensitive text before it is pasted into an AI service, then restores placeholders in the AI response.
For customer support, drafting, development, and other workflows where personal or confidential strings should be removed before AI input.
Base64-only encoding repair API to fix mojibake across UTF-8 / Shift_JIS / EUC-JP / Latin-1 and more.
For legacy exports, migration files, Japanese text recovery, and pre-AI input cleanup.
High-precision multilingual token counter for OpenAI-compatible encodings with a unified JSON response.
For token estimation, cost control, request validation, and logging before LLM execution.
result + meta response schemaPractical CSV cleanup API for escaping, sanitizing, and normalizing structured text before Excel, ETL, or AI ingestion.
For messy CSV exports, spreadsheet safety, ETL cleanup, and LLM-ready structured input.
Utility API for stabilizing messy JSON responses through normalization, transformation, and diff-oriented cleanup.
For unstable third-party APIs, LLM JSON cleanup, and reusable downstream contracts.
We will continue to add utility APIs around preprocessing, quality assurance, and text structuring.
APIron Lab operates this official hub and the products introduced here, including MaskedForm.
For general inquiries, non-sensitive bug reports, and feature requests, use the public GitHub Issues page.
Do not post personal information, credentials, customer logs, unreleased source code, vulnerability details, or other confidential material in a public issue. For confidential security reports, use GitHub Private Vulnerability Reporting.
This is the public support route for APIron Lab products. Paid-service legal disclosures, when required, will be provided separately before sales begin.