Advertise
MCP.so
Sign In

Forge MCP Server

@VoxellInc

Repository

Basic information

Category

Memory & Knowledge

Transports

stdio

Author

VoxellInc

Submitted by

Jonathan Corners

About Forge MCP Server

Forge is Voxell's hosted text-embedding API. This MCP server exposes two tools — embed (turn text into vectors) and list_models — so any MCP-compatible agent can generate embeddings for semantic search, RAG, clustering, and dedup.

#embeddings#text-embedding#vector#semantic-search#rag#retrieval#vector-search#voxell#forge#qwen3#matryoshka#embedding-api#nlp#similarity

Forge MCP Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "forge": {
      "command": "npx",
      "args": [
        "-y",
        "@voxell/forge-mcp"
      ],
      "env": {
        "FORGE_API_KEY": "your-forge-api-key"
      }
    }
  }
}

Forge MCP Server tools

2

Generate vector embeddings for one or more texts with Forge (Voxell's hosted embedding API). Use it to turn text into vectors for semantic search, RAG, clustering, or similarity. Set input_type='query' for search queries and 'document' for content you index. Choose model by quality/cost: turbo (1024d, fast, default) -> pro (2560d) -> ultra (4096d, #4 on MTEB English, top usable). Optionally set dim to truncate (Matryoshka, re-normalized).

List the available Forge embedding models and their dimensions. Call this to pick a model before embedding.

Overview

What is Forge?

Forge is an MCP (Model Context Protocol) server for Voxell's hosted text-embedding API. It exposes two tools (embed and list_models) to any MCP client, including Claude, Cursor, Cline, Windsurf, and VS Code. The server is stateless; Voxell does not store the text sent or the vectors returned — only usage metadata (token counts) for billing.

How to use Forge?

Install via one-click links for Cursor or VS Code, or with a CLI command for Claude Code. Configure your MCP client with a mcpServers block that runs npx -y @voxell/forge-mcp and sets the FORGE_API_KEY environment variable. Requires Node.js ≥ 18 and a Forge API key from https://dash.voxell.ai.

Key features of Forge

  • Provides embed (text to vectors) and list_models tools.
  • Supports Matryoshka (MRL) dimension truncation for smaller vectors.
  • Runs the Qwen3-Embedding model family (turbo, pro, ultra).
  • Low latency (Go + CUDA engine), zero-trust, per-key auth.
  • No storage of text or vectors – only usage metadata recorded.
  • Also offers an OpenAI-compatible API for seamless migration.

Use cases of Forge

  • Add semantic search by embedding documents and queries separately.
  • Build RAG pipelines by embedding a knowledge base and retrieving relevant chunks.
  • Find similar or duplicate text by comparing vector embeddings.
  • Cluster or classify text by embedding batches then applying ML algorithms.
  • Shrink vector storage by truncating dimensions via Matryoshka.

Frequently asked questions

1

What does Forge do vs other MCP servers?

Forge is strictly an embedding server – it does not provide storage, search, or RAG. Those are separate products.

2

What are the runtime requirements?

Node.js version 18 or higher (tested on 20) and a valid Forge API key. New accounts start with 10M free tokens, no credit card required.

3

Where does my data go?

The server is stateless. Voxell does not store the text you send or the vectors it returns. Only usage metadata (token counts) is recorded for billing.

4

What authentication does Forge use?

Authentication is via an API key set in the `FORGE_API_KEY` environment variable. mTLS is available for extra security.

5

What transport does Forge use?

Forge runs as a standard MCP server, launched via `npx`. It connects over standard I/O (stdio) to the MCP client.

Comments

More Memory & Knowledge MCP servers

AOCI-CODE

aoci-spec

Local MCP server and CLI that keeps a Git-versioned, model-written index of a codebase and its database schema, so coding agents start each task with the whole system in view.

1.7K

Takibi Base

Deian Isac

Ask your Takibi Base collections anything; get answers with cited passages and a support score.

Sooveryn

PANDAPROG

Sooveryn gives Claude Code a team of AI personas that remembers your project. Soov, the partner and team lead, keeps your decisions and how you work in a lasting memory, and finds them again by words and by meaning, from

agentleFS

agentleFS

One permissioned folder your team and your AI agents both log into.

UmmahAPI

Yousuf Kazmi

Islamic source data for assistants: Quran, hadith, tafsir, duas, prayer times and qibla. Answers come from real sources with citations instead of recalled text, so verses and hadith are not hallucinated. 36,000+ hadiths

Advertise