Embeddings

Embeddings turn text into numeric vectors that capture meaning, so records with similar content end up close to each other in vector space. This is the model kind behind semantic and similarity search over CRM data — see Vector Search via Solr for the full ADITO Indexsearch integration.

Model

Model ID	adito-embed
Base model	Qwen/Qwen3-Embedding-0.6B
Hosting	ADITO Cloud infrastructure, Germany
Base URL	https://ai.adito.cloud
Endpoint	/embeddings
Interface	OpenAI-compatible Embeddings API
Vector size	1024 dimensions
Context window	32,768 tokens per input string

adito-embed is the floating alias; it currently resolves to the pinned codename adito-embed-andromeda. Naming, the immutability contract and the deprecation lifecycle work the same way here as for every other model kind, see Model naming and versioning.

note

The ADITO Indexsearch #VECTOR field requires vectors with exactly 1024 dimensions. Set the dimensions request parameter accordingly when the resulting vector feeds an Indexsearch vector index.

Default request parameters

Every parameter you do not send is filled in by the runtime with the value below. These defaults are part of what a pinned codename freezes, so they never change for an existing codename — a new codename is issued instead.

Defaults, not limits

A value you send in the request body always wins over the default. Only the context window is a hard limit rather than something you can override.

Defaults for adito-embed-andromeda

Parameter	Default	Description
dimensions	1024	Length of the returned vector. The model supports 32 to 1024 dimensions; Indexsearch #VECTOR fields require exactly 1024.
encoding_format	float	Vectors are returned as an array of floats. Send base64 for the compact representation.
Context window	32768	Hard limit per input string. Longer text must be chunked before embedding. Not a request parameter.

What you can build

• Semantic search — find records by meaning instead of exact keywords, e.g. a search for "contract renewal" also surfacing "subscription extension".
• Similarity search — find related cases, documents, or tickets based on content similarity rather than shared fields.
• Retrieval for prompts — retrieve the most relevant records first, then pass them as context to a Text Generation request. See Building a RAG Use Case for the full pipeline.

First how-to: create an embedding

note

The code examples on this page are not JDito code. They illustrate general API usage and can be adapted to any language or HTTP client. For the JDito integration with ADITO Indexsearch, see Vector Search via Solr.

curl

curl https://ai.adito.cloud/embeddings \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "adito-embed",
    "input": "Customer asked about extending their existing subscription.",
    "dimensions": 1024
  }'

JavaScript

import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'https://ai.adito.cloud',
  apiKey: 'your-api-key'
});

async function getEmbedding() {
  const response = await openai.embeddings.create({
    model: 'adito-embed',
    input: 'Customer asked about extending their existing subscription.',
    dimensions: 1024
  });
  console.log(response.data[0].embedding);
}

getEmbedding();

The response contains one vector per input string, in the same order as the input:

{
  "model": "adito-embed",
  "data": [
    { "index": 0, "object": "embedding", "embedding": [0.027, 0.0419] }
  ],
  "object": "list",
  "usage": { "prompt_tokens": 6, "total_tokens": 6 }
}

---

See also: Vector Search via Solr | Building a RAG Use Case | AI Models