Embeddings — Gamra

POST /v1/embeddings

OpenAI-compatible. Send one string or an array of strings; get one vector per input, in order.

curl https://api.gamra.dev/v1/embeddings \
  -H "Authorization: Bearer $GAMRA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text-embedding-v4",
    "input": ["Livraison en 48 heures", "توصيل في يومين"],
    "dimensions": 1024
  }'

Field
`model`	`text-embedding-v4` — the only embedding model today
`input`	a string or an array of strings
`dimensions`	optional; 1024 is what Gamra Agents uses

Billing is per input token, in millimes, from the same wallet as chat. An embedding model cannot be used with /v1/chat/completions (400 model_not_chat), and a chat model cannot be used here (400 model_not_embedding).