Embeddings API · Production catalog

Turn text into searchable space.

Six models, one OpenAI-compatible endpoint. Compare dimensions, context and cost before indexing your first document.

Create an API key Read the documentation

0: active models
1: endpoint
$0.01: from / 1M tokens

VectorField

Vector footprint

Relative length of the vector returned by each model.

Choose a model

The right vector for every workload.

Prices are per million input tokens. Embeddings do not generate output tokens.

Direct call

Same request, interchangeable model.

Send a string or a list of strings. The response follows the OpenAI embeddings format.

Selected model

BAAI/bge-m3

The response contains a data array with one vector per input and usage information.

curl

curl https://api.routerlab.ch/v1/embeddings \
  -H "Authorization: Bearer $ROUTERLAB_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "BAAI/bge-m3",
    "input": [
      "RouterLab unifie plusieurs modèles derrière une seule API."
    ]
  }'

Ready to index your data?

Create a RouterLab key, keep the same endpoint and switch models without rewriting your integration.

Create an API key