Named Entity Extractor
Find named entities in text and inspect exact, server-calculated occurrence offsets.
Fill in the fields, then submit when you are ready.
Result
Your result will appear here after you submit and complete verification.
How to use Named Entity Extractor
- Paste the source text and optional context about its subject.
- Keep original spelling and spacing if you need offsets for another application.
- Select Extract entities, complete verification and review types and occurrence rows.
Example: Named Entity Extractor
Locate repeated names without rewriting the source.
Options
- Entity types
- The result can label people, organizations, locations, dates, money, email addresses, URLs and other entities. Types remain model judgments.
- Character offsets
- Positions count Unicode code points rather than UTF-8 bytes or JavaScript UTF-16 units. The first position is zero and the end is exclusive.
Supported inputs and limits
- Source text allows 10,000 characters.
- Results contain at most 20 distinct type-and-text pairs and 100 total occurrences.
- Excessive occurrence sets fail rather than being silently cut off.
- Matching is exact and case-sensitive, with overlapping occurrences retained.
Where your input is processed
Your input is uploaded to this site and sent to a configured AI provider. If a provider fails, the request may be sent to the next eligible provider. Check its result before use. Private inputs and results expire within one hour and are excluded from site backups; temporary files are removed by scheduled cleanup. Provider retention rules may differ. Avoid submitting passwords or highly sensitive information.
Keep detection separate from identity
A name appearing in text is evidence of that spelling, not proof of who it describes. Review ambiguous organization names, date expressions and short common words before using the output in a dataset. The site calculates occurrences only after the model selects an exact source string. This makes the positions checkable while leaving the entity selection and type open to human review.
Questions about Named Entity Extractor
Does this identify a real person?
It labels text spans. It does not establish identity, resolve two people with the same name or verify that a name belongs to a particular person.
Will initials and full names be merged?
No automatic identity merge is performed. Distinct spellings remain distinct text values even if they may refer to the same entity.
Can I use offsets in code?
Convert the stated code-point positions if your language or editor uses another index unit. Keep the exact input, including line endings, for reliable slicing.