# anzy-renlab-ai/pronounce [Health: Active]

**Category:** 🎓 Education  
**Repository:** https://github.com/anzy-renlab-ai/pronounce  
**GitHub Stars:** 39  
**Views:** 4  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/anzy-renlab-ai-pronounce

## Description
Community pronunciation dictionary for developer project / product / jargon names (kubectl, nginx, GIF, JSON, Pydantic, Knative, LaTeX — 1,848 entries). Each entry carries IPA, a General-American respelling, alternate readings where the pronunciation is contested, a source citation (creator interview / project FAQ / Wikipedia § Pronunciation), and a confidence level. Tools: pronounce(word), searchpronunciations(query), listpronunciations(category). Install: uvx pronounce-mcp. Live browse: .

## Tools
Capabilities this server exposes over MCP:

- **pronounce** — Look up the canonical pronunciation of a developer-related word.

    Returns the IPA, a respelling that approximates how engineers actually
    say the word, a playable MP3 `audio_url`, optional alternate readings,
    the source citation when available, and editorial notes. Use this for project names like
    `kubectl`, `nginx`, `GIF`, `JSON`, `Pydantic`, `Knative`, programmer
    jargon, and tech acronyms. Returns an error dict if the word is not
    in the dictionary; for general English vocabulary, fall back to your
    own knowledge.
    
- **search_pronunciations** — Search the dictionary by partial word or category match.

    Matches substrings against the entry's `word` and `category` only —
    the search index (/api/words.json) carries no notes or source text, so
    a query like "Kelsey" will NOT surface `kubectl`. To read an entry's
    notes, look it up with `pronounce`.

    Returns a list of matching entries with `word`, `slug`, `category`,
    and `confidence`. Use this when the exact spelling is unknown or
    when browsing entries by category.
    
- **list_pronunciations** — List dictionary entries, optionally filtered by category.

    Categories include: product, project, cli-tool, tool, cs-term,
    acronym, abbreviation, person. The dictionary holds well over 1500
    entries and grows continuously, so the unfiltered list is large —
    pass `category` and/or `limit` to keep the result focused.
    

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "pronounce": {
    "url": "https://api.star-history.com/svg?repos=anzy-renlab-ai/pronounce&type=Date"
  }
}
```

## Documentation

## What anzy-renlab-ai/pronounce MCP server does

The anzy-renlab-ai/pronounce MCP server exposes a community-maintained dictionary of developer-related pronunciations through MCP. Entries cover products, projects, command-line tools, software tools, computer science terms, acronyms, abbreviations, and people. The source material describes a dictionary with more than 1,500 entries, including examples such as kubectl, nginx, GIF, JSON, Pydantic, Knative, and LaTeX.

A pronunciation entry can include International Phonetic Alphabet notation, a General-American-style respelling, an audio URL for an MP3 recording, alternate readings, a confidence level, and a source citation when one is available. Contested terms can therefore represent more than one reading rather than forcing a single answer.

## How it works

The anzy-renlab-ai/pronounce MCP server offers three dictionary operations:

- `pronounce`: Retrieves the detailed entry for an exact word. It returns an error dictionary when the term is absent.
- `search_pronunciations`: Finds entries using a partial word or category match and returns the word, slug, category, and confidence.
- `list_pronunciations`: Lists entries, optionally constrained by category and result limit.

Search only checks the entry’s word and category fields. It does not search notes or source text, so a name mentioned in an editorial note will not necessarily be found by search. Use the exact lookup after identifying a likely entry to obtain the detailed pronunciation data.

## Setup and configuration

Install the MCP server with `uvx pronounce-mcp`. The supplied material does not identify any required environment variables, credentials, API keys, or additional configuration. The repository is licensed under MIT.

The broader project also provides a command-line interface, a website, editor extensions, and other project features, but those are separate from the MCP tools described here. For MCP use, configure the server through the client’s local command setup using the installation command above.

## Tools and capabilities

The dictionary can be browsed by categories including `product`, `project`, `cli-tool`, `tool`, `cs-term`, `acronym`, `abbreviation`, and `person`. Supplying a category or limit is useful because an unrestricted listing can be large.

The detailed lookup is suited to agent responses that need both a readable approximation and a precise IPA form. The returned audio URL can support clients or workflows that can play MP3 content. Citations and confidence fields help distinguish creator-confirmed readings, community conventions, and unresolved alternatives when those fields are present.

## Limitations and notes

The anzy-renlab-ai/pronounce MCP server is focused on developer terminology, not general English pronunciation. For words outside its dictionary, the tool returns an error and the agent should use another source or its own knowledge. Search is not full-text: queries do not match source citations, notes, or other metadata. The material also does not specify a refresh mechanism, remote endpoint, authentication method, or supported MCP client list.

_Full upstream README: https://allmcps.com/mcp/anzy-renlab-ai-pronounce/readme_