Cloud Natural Language API logo

Cloud Natural Language API

Cloud Natural Language API

Google Cloud Natural Language API provides text analysis — entity recognition, sentiment analysis, syntax parsing, content classification (multi-language).

Visit site ↗Health checked 12h ago
Use it when

Multi-language support (10+ major languages including English and Chinese)

Watch for

Generic model — may miss industry-specific terminology

First check

POST /v1/documents:analyzeSentiment for sentiment; analyzeEntities for entities; classifyText for content classification.

Auth
CORS
No
HTTPS
Yes
Signup
?
Protocol
REST
Pricing

Uptime · 30-day window

Probes: 30Uptime: 80%Avg latency: 427ms
01

About this API

Cloud Natural Language API is GCP's long-standing NLP service offering 5 core capabilities: sentiment (sentence and entity-level scoring), entity (people/places/orgs + knowledge-graph linking), syntax (dependency parsing + POS tagging), classification (700+ categories), moderation (inappropriate content detection). In the LLM era this API's positioning is awkward — many tasks can be done via GPT/Gemini prompts, often with higher quality. But Cloud Natural Language still has value in specific scenarios: bulk-task cost (per-character pricing much cheaper than LLMs), low latency (specialized model vs generic LLM), and predictable output (structured JSON vs LLM output variability).

02

What you can build

  • 1Sentiment scoring for user reviews
  • 2Auto-categorize news articles
  • 3Extract entities (people, places, organizations)
  • 4Keyword extraction for content recommendation
03

Strengths & limitations

Strengths

  • Multi-language support (10+ major languages including English and Chinese)
  • Multiple analyses in a single API call
  • High-quality pre-training from Google large models

Limitations

  • Generic model — may miss industry-specific terminology
  • Cannot perform generative tasks (use Vertex AI / Gemini)
04

Getting started

POST /v1/documents:analyzeSentiment for sentiment; analyzeEntities for entities; classifyText for content classification.

05

FAQ

Is Chinese sentiment analysis accurate?+

Decent for everyday text (reviews, social media). For professional domains (finance, healthcare), recommend custom training.

How does it compare to Gemini?+

High volume + simple task + cost-sensitive: Language API. Complex understanding, context, generation: Gemini.

06

Technical details

CORS: NoHTTPS: YesSignup: ?Open source: No
Auth type
unknown
Pricing
unknown
Protocols
REST
SDKs
python, javascript, go, java, csharp
Last health check
6/26/2026, 6:23:28 AM
07

Tags

08

More from Google