Integration

Babel Street Analytics integration

Babel Street Analytics (formerly Rosette Text Analytics) is a multilingual NLP API. Match and score names, addresses and whole records across languages and scripts, translate names, deduplicate name lists, and analyze text: extract entities, events, relationships, topics and categories, detect language, score sentiment, tokenize, split sentences, run morphology and syntax, transliterate Arabizi, and get text embeddings or similar terms. Each analysis has a supported-languages lookup. The on-premise-only gazetteer route is not exposed. Cloud limits: 600 KB or 50,000 characters per call, one call at a time per key. Authentication is an API key from the Analytics Developer Portal sent in the X-BabelStreetAPI-Key header.

Connect Babel Street Analytics How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Babel Street Analytics.Built from the 43 Babel Street Analytics tools below, on the connection you add once.

What you can do with Babel Street Analytics

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Supported language, 20 operations

list_address_similarity_languagesList address similarity languagesGETList the language pairs match_addresses supports: source and target language, script and transliteration scheme, and whether your key is licensed for each pair.
list_category_languagesList category languagesGETList the languages and scripts categorize_text supports, and whether your key is licensed for each.
list_entity_coreference_languagesList entity coreference languagesGETList the languages and scripts extract_entities with options.useIndocServer (in-document coreference) supports, and whether your key is licensed for each.
list_entity_languagesList entity languagesGETList the languages and scripts extract_entities supports, and whether your key is licensed for each.
list_event_languagesList event languagesGETList the languages and scripts extract_events supports, and whether your key is licensed for each.
list_language_detection_languagesList language detection languagesGETList the languages and scripts detect_language supports, and whether your key is licensed for each.
list_morphology_languagesList morphology languagesGETList the languages and scripts analyze_morphology supports, and whether your key is licensed for each.
list_name_deduplication_languagesList name deduplication languagesGETList the languages and scripts deduplicate_names supports, and whether your key is licensed for each.
list_name_similarity_languagesList name similarity languagesGETList the language pairs match_names supports: source and target language, script and transliteration scheme, and whether your key is licensed for each pair.
list_name_translation_languagesList name translation languagesGETList the language pairs translate_name supports: source and target language, script and transliteration scheme, and whether your key is licensed for each pair.
list_record_similarity_languagesList record similarity languagesGETList the language pairs match_records supports: source and target language, script and transliteration scheme, and whether your key is licensed for each pair.
list_relationship_languagesList relationship languagesGETList the languages and scripts extract_relationships supports, and whether your key is licensed for each.
list_sentence_languagesList sentence languagesGETList the languages and scripts split_sentences supports, and whether your key is licensed for each.
list_sentiment_languagesList sentiment languagesGETList the languages and scripts analyze_sentiment supports, and whether your key is licensed for each.
list_similar_terms_languagesList similar terms languagesGETList the languages and scripts find_similar_terms supports, and whether your key is licensed for each.
list_syntax_dependency_languagesList syntax dependency languagesGETList the languages and scripts parse_syntax_dependencies supports, and whether your key is licensed for each.
list_text_embedding_languagesList text embedding languagesGETList the languages and scripts get_text_embedding supports, and whether your key is licensed for each.
list_tokenization_languagesList tokenization languagesGETList the languages and scripts tokenize_text supports, and whether your key is licensed for each.
list_topic_languagesList topic languagesGETList the languages and scripts extract_topics supports, and whether your key is licensed for each.
list_transliteration_languagesList transliteration languagesGETList the languages and scripts transliterate_arabic supports, and whether your key is licensed for each.

Name, 3 operations

deduplicate_namesDeduplicate namesPOSTGroup a list of names into clusters of likely duplicates. Returns results: one cluster id per input name, in input order; names sharing a cluster id are considered duplicates.
match_namesMatch namesPOSTScore how likely two names refer to the same entity, across languages and scripts (Michael Jackson vs a Chinese spelling). Returns score, from 0 to 1.
translate_nameTranslate namePOSTTranslate or transliterate a proper name into a target language. Returns translations with confidence (one by default; set maximumResults for more candidates) and the detected source script and language.

Address, 1 operation

match_addressesMatch addressesPOSTScore how likely two addresses are the same place, comparing their fields. Returns score, from 0 to 1.

Category, 1 operation

categorize_textCategorize textPOSTClassify a text (content or contentUri) into Tier 1 contextual categories of the QAG taxonomy. Returns categories with label, confidence and score.

Dependency, 1 operation

parse_syntax_dependenciesParse syntax dependenciesPOSTParse a text (content or contentUri) into syntactic dependencies: for each sentence, labeled directed links between tokens (by token index), plus the list of tokens.

Embedding, 1 operation

get_text_embeddingGet text embeddingPOSTCompute the text embedding of a text (content or contentUri): one vector of floating point numbers, usable to compare the similarity of two documents or words. Per-token embeddings on request.

Entity, 1 operation

extract_entitiesExtract entitiesPOSTExtract named entities (people, places, organizations...) from a text (content or contentUri) and link them to a knowledge base. Returns entities with type, mention, normalized form, count, offsets and entityId (Wikidata QID or temporary id).

Entity type, 1 operation

list_entity_typesList entity typesGETList the entity types extract_entities can return, per language and processor (and per subsource when perSubsourceSupportedEntityTypes is true).

Event, 1 operation

extract_eventsExtract eventsPOSTExtract events from a text (content or contentUri). Returns each event with its type, mentions (offsets, roles, polarity, negation cues), confidence and the workspaceId of the event model that found it.

Event schema, 1 operation

list_event_schemasList event schemasGETList the event models (workspaces) available to extract_events and their schemas: event types, key phrases, roles, role types and extractors. Give workspaceId for one model.

Language, 1 operation

detect_languageDetect languagePOSTIdentify the language of a text (content or contentUri). Returns candidate languages with confidence, highest first. For the language of a NAME, use match_names or translate_name instead (tuned for names).

Morphology, 1 operation

analyze_morphologyAnalyze morphologyPOSTRun morphological analysis on a text (content or contentUri): lemmas, part-of-speech tags, compound components or Han readings, chosen by morphoFeature. Which features are returned depends on the language.

Record, 1 operation

match_recordsMatch recordsPOSTScore pairs of records (people or organizations with names, dates, addresses, ids) field by field. Returns results, each with a score, the two records and, on request, explainInfo.

Relationship, 1 operation

extract_relationshipsExtract relationshipsPOSTExtract relationships between entities in a text (content or contentUri). Each relationship has a predicate (usually the main verb), arg1 (usually the subject), optional arg2 (the object), optional knowledge-base ids and a confidence between 0 and 1.

Sentence, 1 operation

split_sentencesSplit sentencesPOSTSplit a text (content or contentUri) into sentences. Returns the list of sentences.

Sentiment, 1 operation

analyze_sentimentAnalyze sentimentPOSTScore the sentiment of a text (content or contentUri). Returns the document sentiment label with a confidence, and the entities found in it.

Service info, 1 operation

get_service_infoGet service infoGETGet the name, version, build number and build time of the Babel Street Analytics service.

Service status, 1 operation

pingPingGETCheck that Babel Street Analytics is reachable and the API key is accepted. Returns a message and the server time.

Similar term, 1 operation

find_similar_termsFind similar termsPOSTFind terms semantically similar to the input term or text (content), optionally in other languages. Returns, per requested language, a list of terms with a similarity score.

Token, 1 operation

tokenize_textTokenize textPOSTSplit a text (content or contentUri) into tokens. Returns the list of tokens.

Topic, 1 operation

extract_topicsExtract topicsPOSTExtract the topics of a text (content or contentUri): key phrases and concepts, each with a salience score.

Transliteration, 1 operation

transliterate_arabicTransliterate arabicPOSTTransliterate Arabic written in Arabizi (romanized chat Arabic) into native Arabic script, or the reverse with options.reversed. Returns the transliterated string.

Use Babel Street Analytics from Claude Code, Cursor or Codex

Connect Babel Street Analytics in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Babel Street Analytics operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Babel Street Analytics

Create your workspace
Sign up free. Nothing to install.
Connect Babel Street Analytics once
Add your Babel Street Analytics connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.