Named Entity Extraction & Sentiment Analysis API

Turn unstructured text into clean, ready-to-use intelligence. Paste comments, reviews, support tickets, social posts, transcripts, survey answers, news snippets, or a CSV/Excel file and get a structured result for every row: sentiment, named entities, or both.

Built for teams who need reliable named entity extraction, sentiment analysis, brand monitoring, and text classification at scale—without manual tagging, spreadsheet clean-up, or accounts to connect.

What you get

• One result per text — every input line or file row is returned in the Output tab as soon as it is ready.
• Clear sentiment — positive, neutral, or negative.
• Useful named entities — up to 50 relevant people, organizations, brands, products, places, events, and other specific references per text.
• Original text preserved — exports always include the exact text that was analyzed. File-based runs also preserve the original file row for easy joining.
• Bulk-friendly workflow — add texts one per row or upload a single-sheet CSV, XLS, or XLSX file.
• CSV, Excel, JSON, and API ready — use the dataset directly in a dashboard, workflow, CRM, or AI agent.

Who this is for

• Social listening and reputation teams
• Brand, product, and customer-experience teams
• Market researchers and analysts
• Agencies managing multi-brand campaigns
• Support, success, and VOC teams
• Newsrooms, researchers, and intelligence teams
• Sales and operations teams enriching text-heavy data
• Builders connecting Apify to Make, Zapier, n8n, or an AI agent

Popular use cases

Social media listening and brand intelligence

Paste social posts, comments, community discussions, or video transcriptions to identify brand names, products, people, locations, and campaign themes while measuring the tone around each mention. Use it to spot praise, complaints, competitor mentions, creators, and emerging conversations.

Customer review analysis

Upload a review export and immediately see the sentiment of every review plus the products, branches, people, or services customers mention most often. Export the dataset to prioritize response queues and identify recurring issues.

Voice of customer and survey analysis

Classify open-ended survey answers, NPS feedback, chat transcripts, and support tickets. Surface references to features, teams, plans, regions, and competitors without forcing respondents into rigid form fields.

PR and news monitoring

Analyze articles, press coverage, newsletters, and media summaries in bulk. Capture companies, executives, products, events, and places, then use the sentiment label as a quick direction-of-coverage signal.

Podcast, video, and meeting transcript intelligence

Feed in a transcript from an interview, webinar, call, or video. Find the people, brands, products, and places that matter, then use the output to create clips, show notes, follow-ups, or research lists.

App, marketplace, and e-commerce feedback

Turn raw product feedback into a usable table. Identify which feature, SKU, delivery provider, seller, or location each comment refers to, and compare how customers feel across the data.

Input

Choose what to extract, then use one of two simple ways to supply text:

Field	What to do
What would you like to extract?	Select Sentiment, Named entities, or both.
Texts to analyze	Add one text per row with the bulk input. Each row accepts up to 2,500 characters.
CSV or Excel file	Upload a CSV, XLS, or XLSX file with one sheet for a larger batch.
Text column name	For an uploaded file, enter the header of the column that contains the text.
Convert entity names to lowercase	Optional—helpful for matching and deduplicating names across exports.

Example: pasted text

{
  "analysisType": "both",
  "texts": ["I love the new Acme Cloud product from Milan.", "The delivery from Northwind was disappointing."],
  "lowercaseEntities": true
}

Example: file analysis

{
  "analysisType": "both",
  "inputFile": "https://example.com/customer-feedback.xlsx",
  "textColumn": "review_text",
  "lowercaseEntities": false
}

Output

The default dataset contains one clean record per text. In the Apify Output tab, use Overview for the analysis table or Original file rows to keep file columns beside the classifications.

Field	Meaning
originalText	The original text submitted for analysis.
sentiment	positive, neutral, or negative; empty when sentiment was not selected.
entities	An array of up to 50 relevant named entities, or null when a file row has no text.
entityCount	Number of returned entities, or null when a file row has no text.
error	A user-readable message only if that text could not be processed.

Example output

{
  "rowNumber": 1,
  "originalText": "I love the new Acme Cloud product from Milan.",
  "sentiment": "positive",
  "entities": ["acme cloud", "milan"],
  "entityCount": 2,
  "error": null
}

Connect it to your workflow

• Make or Zapier: send new reviews, form responses, tickets, or social posts to the Actor, then write sentiment and entities back to your sheet, CRM, or alerting tool.
• n8n: run scheduled batches from a database or spreadsheet, then route negative feedback to a support queue and entity-rich records to a research workflow.
• Google Sheets or Excel: export the dataset after every run for a clean, filterable analysis table.
• AI agents and MCP: the Actor’s clear input and output schemas make it easy for Claude, Codex, Hermes, and other agent workflows to understand the available inputs and consume structured results.
• Apify API: start a run with JSON input and fetch the default dataset whenever your workflow is ready.

Discover more Lofomachines Actors

Build a complete listening and intelligence workflow with these related Actors:

• Sentiment and Topics Text Classifier
• Fast Audio and Video Transcription
• TikTok Transcription AI
• Brand Monitoring AI
• AI Media Monitor

Explore all Lofomachines Actors →