MonkeyLearn MCP. Analyze text sentiment and extract keywords instantly.

You connect your AI client to this server and get immediate access to MonkeyLearn's NLP models. You don't need a data scientist or boilerplate code; you just send raw text, and we do the heavy lifting for classifying sentiment, pulling out specific facts, and finding keywords.

Classifying Sentiment and Topics

The classify_text tool runs any given block of text through one of your configured NLP models. You get back a structured result: not just 'Positive' or 'Negative,' but also a confidence score, telling you how sure the model is about its guess. If you need to know what topics are even available for classification, run list_classifiers to see every sentiment or topic detection model set up in your account.

You can then drill down into exactly what any single classifier expects as input using get_classifier_details(model_id).

Extracting Specific Data Points

The extract_text tool scans unstructured text and pulls out structured information. This means if you feed it a customer review, it doesn't just tell you the sentiment; it can pull specific names, product IDs, dates, or keywords based on an extractor model you defined. To see what kinds of data extraction models are available for your use, call list_extractors.

If you need to know the precise requirements—like whether the text needs a date format or if it only accepts simple strings—you check the metadata using get_extractor_details(model_id).

Understanding Your NLP Ecosystem
The server gives you full visibility into your entire setup. You can run list_pipelines to see automated workflows that combine multiple steps—for example, one workflow might automatically classify sentiment and pull out all names in the same go. Similarly, list_workflows shows all these pre-built data processing jobs ready to fire up when you call them.

To understand the underlying structure of your classification models, run list_tag_trees. This reveals the organizational hierarchy—the taxonomy—that dictates how those topics are nested and categorized. The toolset also lets you check what's going on with all automated processes by calling list_pipelines or list_workflows; it’s how you keep track of your full processing capability.

Tracking Usage and Activity

You don't have to guess what's happening behind the scenes. The list_activity tool shows you a complete history of your account's processing activity, telling you exactly when calls were made and how many times you ran specific models over time. This is crucial for monitoring usage.

In Short: You send text to the server, it uses the appropriate model—whether that's classifying sentiment with classify_text or pulling out product IDs with extract_text—and gives you clean, structured data back every single time.