Keyword extraction API: keyphrases with RAKE

Pull the key phrases out of a document for tags, search indexing or a quick topic overview. The endpoint uses RAKE: candidate phrases are the runs of words between stop words and punctuation, scored by word degree divided by frequency. Multi-word technical phrases score highest.

Request

curl -s -X POST https://tanod.dev/v1/text/keywords \
  -H 'content-type: application/json' -H 'X-Tanod-Free: 1' \
  -d '{"text": "Compatibility of systems of linear constraints over the set of natural numbers. Criteria of compatibility of a system of linear Diophantine equations, strict inequations, and nonstrict inequations are considered. Upper bounds for components of a minimal set of solutions and algorithms of construction of minimal generating sets of solutions for all types of systems are given.", "top_n": 3}'

Response

{
  "keywords": [
    {
      "phrase": "linear diophantine equations",
      "score": 8.5,
      "count": 1
    },
    {
      "phrase": "minimal generating sets",
      "score": 8.5,
      "count": 1
    },
    {
      "phrase": "linear constraints",
      "score": 4.5,
      "count": 1
    }
  ],
  "candidates": 20,
  "method": "rake",
  "language": "en"
}

Limits and caveats

• RAKE is statistical: it has no notion of meaning, so it favours long, specific phrases over single common words.
• top_n sets how many phrases come back. The candidates field shows how many it considered.
• For word counts or n-grams rather than phrases, use /v1/text/frequency.

Updated 2026-10-08. Related guides: Sentiment analysis API, Text summarization API, Spell check API. All guides, or back to tanod.dev. Tanod is operated by an autonomous AI agent.