Bareun — Korean NLP & Spell/Grammar Checking MCP
Official/What is Bareun — Korean NLP & Spell/Grammar Checking MCP?
Bareun is a Korean natural-language platform exposed as an MCP server, letting MCP-compatible AI tools such as Claude, Cursor, VS Code and Claude Desktop perform Korean morphological analysis and spell/grammar correction by calling Bareun as a tool. It is hosted at api.bareun.ai/mcp over Streamable HTTP, and the same endpoint works on self-hosted installs.
Large language models still miss subtle Korean spacing, particle agreement and confusable-word rules; plugging Bareun in lets an agent hand off analysis and proofreading to a dedicated Korean engine and use the result to produce more accurate Korean output. Corrections come back as blocks with categories and rules, so the agent can explain each edit rather than silently rewriting the sentence.
The homograph sense disambiguation feature (with_sense) is in beta and ships officially with Bareun 3.1.0.
AI-generated listing based on the project's public information. It may contain errors: check the original source.
/What you can do with Bareun — Korean NLP & Spell/Grammar Checking MCP
/Tools Bareun — Korean NLP & Spell/Grammar Checking MCP gives your assistant
What your assistant can call once connected.
| Tool | What it does |
|---|---|
| analyze_syntax | Performs Korean morphological analysis: splits a sentence into words and morphemes with part-of-speech tags. Supports compact or full output and optional homograph sense disambiguation (beta). |
| analyze_syntax_raw | Performs morphological analysis without post-processing: no compound splitting, auto spacing or custom dictionaries, returning the raw model output for evaluation or corpus retagging. |
| correct_grammar | Corrects Korean spelling and spacing, returning the corrected sentence plus correction blocks with rule categories and explanations, without rewriting the original style. |
| list_pos_tags | Returns the 47 part-of-speech tags Bareun uses (code, name, class) as JSON, for interpreting POS codes in analyze_syntax results. |
| search_dict | Searches the Urimalsaem Korean dictionary (~1.1M entries) by jamo (phoneme-level) slot patterns, e.g. verbs whose stem ends in the ㅎ coda or words ending in -아지. |
| tokenize | Splits a Korean sentence into word (token) units without part-of-speech information; with auto_spacing it also fixes broken spacing before splitting. |
/Questions about Bareun — Korean NLP & Spell/Grammar Checking MCP
Do I need a key or an account?
No. It connects without a key.
Which assistants does it work with?
Claude, ChatGPT, Gemini, Grok
Which countries does it work in?
Any country.
