canonical: https://jentic.com/apis/api.cloudmersive.com/cloudmersive

# Cloudmersive Natural Language Processing API

Jentic publishes the only available OpenAPI specification for Cloudmersive Natural Language Processing API, keeping it validated and agent-ready. Cloudmersive's NLP v2 API provides ready-made text analysis endpoints - sentiment analysis, profanity and hate-speech detection, subjectivity scoring, semantic similarity, named-entity extraction, language detection, machine translation between English and several major languages, parsing, part-of-speech tagging, rephrasing, segmentation, and spell checking. Each endpoint is a simple POST that accepts a JSON body of text and returns a structured result, with no model training or hosting required by the caller.

## For AI agents

Run sentiment, profanity, hate-speech, similarity, entity-extraction, language-detection, translation, and spell-check operations on text via simple POST endpoints.

## Scope

Does not handle large language model chat completion, embeddings, or speech-to-text - use for fixed-purpose NLP scoring, translation, and text-cleanup endpoints only.

## Capabilities

- Score sentiment, profanity, hate speech, and subjectivity for arbitrary input text
- Extract named entities (people, organisations, places) and detect language from text
- Translate text between English and German, French, Russian, and other major languages
- Compare two strings for semantic similarity to power deduplication or matching workflows
- Run grammar and spelling checks plus rephrasing and part-of-speech tagging on input text

## Use cases

### User-Generated Content Moderation

Filter user comments, reviews, or chat messages for profanity and hate speech before they reach other users. The agent calls POST /nlp-v2/analytics/profanity and POST /nlp-v2/analytics/hate-speech with the message body and inspects the boolean and score fields. Combine with a sentiment call to flag unusually negative content for human review.

Example prompt: Call POST /nlp-v2/analytics/profanity with a comment body and reject the post when the returned score exceeds 0.5.

### Multilingual Customer Support Triage

When tickets arrive in mixed languages, detect each ticket's language and translate it to the support team's working language. The agent first calls POST /nlp-v2/language/detect, then routes English/German/French content to the matching translate endpoint such as /nlp-v2/translate/language/deu/to/eng. Sentiment scoring on the translated text helps escalate angry customers first.

Example prompt: Call POST /nlp-v2/language/detect on the ticket body, then POST /nlp-v2/translate/language/{src}/to/eng to translate it before assigning to the English queue.

### Review Sentiment Aggregation

Score customer reviews in bulk to produce a sentiment trend by product or service. The agent posts each review to /nlp-v2/analytics/sentiment, stores the score and label, and rolls up the daily averages into a dashboard. Pair with /nlp-v2/extract-entities to break out which features are mentioned most often.

Example prompt: Call POST /nlp-v2/analytics/sentiment for each new review and write the score plus product ID to BigQuery.

### Agent-Driven Text Analysis via Jentic

Connect Cloudmersive NLP to an AI agent through Jentic so a workflow can ask 'is this comment offensive?' or 'what language is this?' as a single tool call. The agent searches Jentic for the right Cloudmersive operation, loads the schema, and posts the text with the API key kept in your Jentic One instance. The Apikey header value never enters the agent's prompt.

Example prompt: Use the Jentic Python SDK to search for 'detect profanity in text', load POST /nlp-v2/analytics/profanity, and execute it on a candidate comment.

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | /nlp-v2/analytics/sentiment | Analyse sentiment of text |
| POST | /nlp-v2/analytics/profanity | Detect profanity in text |
| POST | /nlp-v2/analytics/hate-speech | Detect hate speech in text |
| POST | /nlp-v2/analytics/similarity | Compare semantic similarity of two texts |
| POST | /nlp-v2/extract-entities | Extract named entities from text |
| POST | /nlp-v2/language/detect | Detect the language of input text |
| POST | /nlp-v2/translate/language/eng/to/deu | Translate English to German |

## Key resources

- **Analytics** — Sentiment, profanity, hate-speech, subjectivity, and similarity scoring.
- **Language** — Detect input language and translate between supported language pairs.
- **Entities** — Extract named entities such as people, places, and organisations.
- **Text** — Parsing, part-of-speech tagging, segmentation, rephrasing, and spell checking.

## Why Jentic

- **Setup:** Wiring the Cloudmersive NLP API by hand means handling its Apikey header and mapping the sentiment, entity, language, and translation routes yourself. Through Jentic you install once, import Cloudmersive NLP from the API Directory, store the key once, and your agent calls it.
- **Permission scoping:** Cloudmersive NLP addresses each analysis through the request body rather than resource ids in the URL path, so you limit the agent to the operations it needs, such as sentiment scoring or language detection. Operations like translation are separate and are only available if you include them in the allowed set.
- **Credential handling:** Your Cloudmersive Apikey is stored once, encrypted, by your own Jentic One instance and injected as the header at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'detect profanity' or 'translate text', and Jentic returns the matching Cloudmersive operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

## Related APIs

- **OpenAI** — GPT-class chat-completion API that can perform sentiment, classification, and translation via prompts.
- **Cohere** — Hosted classification, embedding, and generate APIs aimed at developers building NLP workflows.
- **Microsoft Translator** — Wider language coverage and document-level translation than Cloudmersive's pairwise endpoints.

## FAQ

### Why is there no official OpenAPI spec for Cloudmersive Natural Language Processing API?

Cloudmersive publishes interactive HTML reference docs but does not host a single canonical OpenAPI specification for the NLP v2 product. Jentic generates and maintains this spec so that AI agents and developers can call Cloudmersive Natural Language Processing API via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

### What authentication does the Cloudmersive NLP API use?

The API uses an API key in the Apikey header on every request. You generate the key from the Cloudmersive account dashboard. Through Jentic, the Apikey value is stored encrypted in the vault and only scoped tokens reach the agent's context.

### Can I detect hate speech in text with Cloudmersive NLP?

Yes. POST the text to /nlp-v2/analytics/hate-speech and Cloudmersive returns a hate-speech probability and a boolean classification. Pair it with /nlp-v2/analytics/profanity for layered comment moderation.

### What are the rate limits for the Cloudmersive NLP API?

Cloudmersive applies a per-account monthly quota that depends on your plan; free tiers cap at a few hundred calls per month, paid tiers scale into the millions. Per-second throttling is in place to prevent burst overload - when running through Jentic, watch 429 responses.

### How do I translate text between English and German via Jentic?

Search Jentic for 'translate text english to german' and the POST /nlp-v2/translate/language/eng/to/deu Cloudmersive operation is returned with its body schema. Load the operation, then execute it with the source string. The standard quickstart is pip install jentic, search, load, execute.

### Is the Cloudmersive NLP API free?

Cloudmersive offers a free tier with a monthly call quota suitable for evaluation; paid plans raise the quota and unlock higher concurrency. There is no trial expiry on the free tier as long as you stay within the monthly cap.

### Can I limit what my agent is allowed to do with the Cloudmersive NLP API?

Yes. Because Cloudmersive NLP selects each analysis through the request body rather than resource ids in the URL path, your self-hosted Jentic One instance lets you allow only the operations the agent actually needs, such as sentiment scoring at /nlp-v2/analytics/sentiment or language detection at /nlp-v2/language/detect. Separate operations like translation at /nlp-v2/translate/language/eng/to/deu are reachable only when you add them to the allowed set. Your own rules decide which endpoints and which Apikey credential the agent may use, so it cannot call anything outside that scope.