mcp-embeddings mcp

v1.0.0 · MCP Tool · ai · registry.pascalai.org

Generate text embeddings using any configured LLM provider via MakerAI (LLM_Embed / LLM_EmbedBatch). Returns dense vector representations suitable for semantic search, clustering and RAG pipelines. Includes cosine similarity and euclidean distance helpers.

embeddingsvectorssimilaritysemanticrag

MakerAI Pipeline

Input Parameters

ParameterTypeDescription
operationrequired string 'embed': single text; 'embed_batch': multiple texts; 'similarity': cosine similarity between two texts; 'distance': euclidean distance. One of: embed, embed_batch, similarity, distance.
textoptional string Text to embed (for 'embed' and one side of 'similarity'/'distance').
textsoptional array[string] List of texts for batch embedding.
text_boptional string Second text for 'similarity' or 'distance' operations.
modeloptional string Embedding model. Examples: 'text-embedding-3-small', 'text-embedding-3-large', 'embed-english-v3.0'. Empty = provider default.
provideroptional string One of: openai, cohere, google, ollama, auto. Default: auto.

Output Fields

FieldTypeDescription
operation string
vector array[number] Embedding vector (single embed).
vectors array[array] List of embedding vectors (batch).
dimensions integer Vector dimensionality.
model string
tokens_used integer
similarity number Cosine similarity score [-1, 1] (similarity operation).
distance number Euclidean distance (distance operation).

Examples

Embed a single text

// Input
{
  "operation": "embed",
  "text": "The quick brown fox jumps over the lazy dog",
  "model": "text-embedding-3-small"
}

// Output
{
  "operation": "embed",
  "vector": [
    0.0234,
    -0.1823,
    0.0571,
    "..."
  ],
  "dimensions": 1536,
  "model": "text-embedding-3-small",
  "tokens_used": 10
}

Compute semantic similarity between two sentences

// Input
{
  "operation": "similarity",
  "text": "How do I install Python?",
  "text_b": "Python installation guide"
}

// Output
{
  "operation": "similarity",
  "similarity": 0.923,
  "model": "text-embedding-3-small"
}

Install & Discovery

Install

ppm install mcp-embeddings

Get JSON Schema

GET /v1/packages/mcp-embeddings/1.0.0/schema

Discover by keyword

GET /v1/mcp/discover?q=embeddings
Discovery hint: Install with ppm install mcp-embeddings or invoke remotely via POST /v1/invoke/mcp-embeddings on the MCP Service.

PascalAI Usage

uses toolslib;
var Tool := LoadTool('mcp-embeddings');
var R := Tool.Call(JsonObj(['operation','similarity','text','cat','text_b','kitten']));
Writeln(FloatToStr(R['similarity']));