ESC

Text Analysis Toolkit

Run deterministic text analysis across 9 endpoints. Compute readability scores, produce extractive summaries, compare documents, create structured diffs, mask sensitive data, filter profanity, analyze term frequency, detect language, and transliterate Unicode text.

Base URL

https://textanalysis.toolkitapi.io/v1/

Endpoints

Readability and Summarization

Method Endpoint Description
POST /v1/text/readability Compute readability metrics and audience interpretation
POST /v1/text/summarize Build an extractive summary with top ranked sentences

Similarity and Diff

Method Endpoint Description
POST /v1/text/similarity Compare two strings using Levenshtein, cosine, or Jaccard
POST /v1/text/diff Generate unified and structured line-level differences

Filtering and Frequency

Method Endpoint Description
POST /v1/text/pii-mask Detect and mask PII types such as email, phone, and SSN
POST /v1/text/profanity Detect and optionally mask profane terms
POST /v1/text/word-frequency Return top word frequencies with percentages

Language Tools

Method Endpoint Description
POST /v1/text/language Detect language and return ranked candidates
POST /v1/text/transliterate Convert Unicode text to ASCII approximation

Quick SDK Example

from toolkitapi import TextAnalysis

with TextAnalysis(api_key="tk_...") as ta:
    readability = ta.readability(text="This is a short sample paragraph for scoring.")
    similarity = ta.similarity(
        a="The quick brown fox jumps over the lazy dog.",
        b="A quick brown fox jumped over a lazy dog.",
        method="cosine",
    )

print(readability["scores"]["flesch_reading_ease"])
print(similarity["similarity"])

Python SDK

pip install toolkitapi

from toolkitapi import TextAnalysis

with TextAnalysis(api_key="tk_...") as ta:
    result = ta.language(text="Bonjour tout le monde", top_n=3)
    print(result["detected"], result["confidence"])

See drilldowns for endpoint-specific request and response fields.

Tip

Language detection requires at least 10 characters for reliable scoring. For very short inputs, aggregate adjacent text before calling the endpoint.

Previous
Mock Endpoints
Next
Readability and Summarization