Which encoding options are supported?

Encoding can be set to utf32, utf16, or utf8 for morpheme offset calculations.

Can I use custom dictionaries?

Yes, supply custom_dict_names in analyze_syntax or correct_grammar calls.

Bareun — Korean NLP & Spell/Grammar Checking

Verified

Korean NLP server for morphological analysis and spell/grammar correction.

MCP ServerOtherRemote (streamable-http)

View on GitHub

Updated 2026-06-15

What is the Bareun — Korean NLP & Spell/Grammar Checking MCP server?

Bareun exposes four tools over the hosted endpoint https://api.bareun.ai/mcp using Streamable HTTP transport and API-key authentication. Clients can perform precise Korean morphological tagging, tokenization, and grammar correction without local installation.

The server also publishes resources for POS tags, server metadata, and custom dictionaries, all protected by the same API key. It is designed for MCP-compatible environments such as Claude, Cursor, and VS Code.

Install & connect

Add this to your MCP client config. Pick your client below and copy.

{
  "mcpServers": {
    "bareun": {
      "url": "https://api.bareun.ai/mcp"
    }
  }
}

Tools it exposes

4 tools your AI client can call once connected.

analyze_syntax— Splits a sentence into words/morphemes and tags parts of speech (morphological analysis).

textauto_split_sentenceauto_spacingauto_jointingcustom_dict_namesencodingformat

tokenize— Splits a sentence into word (token) units.

textauto_spacingencoding

correct_grammar— Corrects spelling/spacing and returns correction blocks.

textcustom_dict_namestreat_as_titledisable_split_sentencedisable_caret_spacingdisable_vx_spacingenable_limited_punctuationdisable_confusionenable_cleanup_whitespacedisable_typo_correction

list_pos_tags— Returns the 47 part-of-speech tags Bareun uses (code · name · class).

Example prompts

Once connected, try asking your AI client:

“Analyze the syntax of this Korean sentence: 나는 학교에 간다.

“Correct the grammar and spelling in the following Korean paragraph.

“Tokenize this Korean text and list the resulting tokens.

“What are the POS tags that Bareun recognizes?

Security & permissions

Requires a valid Bareun API key passed via the api-key header or Authorization Bearer token; all requests are sent to the remote endpoint https://api.bareun.ai/mcp.

What you can do with Bareun — Korean NLP & Spell/Grammar Checking

Accurate Korean text analysis

Use analyze_syntax to break down sentences into morphemes with POS tags for downstream processing.

Grammar and spelling correction

Call correct_grammar on Korean text to receive spelling, spacing, and particle corrections.

Tokenization for NLP pipelines

Tokenize Korean sentences into word units before feeding them into machine-learning models.

How to use Bareun — Korean NLP & Spell/Grammar Checking

1Obtain an API key from https://bareun.ai.
2Register the server in your MCP client using transport http and the URL https://api.bareun.ai/mcp.
3Add the header api-key: YOUR_API_KEY in the client configuration.
4Restart the client or reload the MCP configuration.
5Test connectivity with the MCP Inspector or by calling list_pos_tags.