Skip to content
Sentiment analysis · REST API

Understand how your users feel — in one API call.

Send text, get back a sentiment label with a confidence score. Built for product teams, support tooling, and analytics pipelines.

Classes
4
Languages
12
p99 target
<500ms

01 — How it works

Text in, sentiment out.

One endpoint POST /v1/analyze. No SDK lock-in.

  1. 01

    Send your text

    Up to 5,000 characters per request, in any of 12 languages.

  2. 02

    We score it live

    Sentiment is analyzed in real time and returned synchronously.

  3. 03

    Get a clean result

    A single label plus a confidence score from 0.0 to 1.0.

02 — Features

Built for production sentiment.

Four classes, each with confidence

Every request returns one label and a confidence score from 0.0 to 1.0.

POSITIVE NEGATIVE NEUTRAL MIXED

12 languages, one schema

English, Spanish, French, German, Italian, Portuguese, Arabic, Hindi, Japanese, Korean, Simplified Chinese, and Traditional Chinese.

Real-time scoring

Synchronous results — no batch jobs, no polling.

Up to 5,000 chars

A one-line review or a long thread, in one request.

Secure API key + token auth

A secure API key plus a JWT bearer token on every request.

Low latency, metered simply

Real-time responses with a sub-500ms p99 target, metered per request and billed monthly.

p99 < 500ms
03 — Live demo

The Sentiment Compass.

Pick a sample and watch the needle find the sentiment. Runs entirely in your browser.

Analyzing

“I absolutely love this product!”

language: en

Try a sample

Illustrative only — no live network call.

12 supported languages

04 — Use cases

Drop it into what you already run.

Product reviews

Track how sentiment trends release over release.

Support tickets

Auto-route and prioritize by customer tone.

Social monitoring

Watch brand sentiment across channels in real time.

Survey & NPS

Turn open-ended feedback into scored signal.

05 — Contact

Tell us about your use case.

Reach out and we'll get you API access. Tell us what you're building and how sentiment fits in.

Prefer email

soheil.ai.compassx@gmail.com
What do you want to analyze?