> ## Documentation Index
> Fetch the complete documentation index at: https://docs.meliai.app/llms.txt
> Use this file to discover all available pages before exploring further.

# POST /v1/embeddings — Generate Vector Embeddings

> POST /v1/embeddings converts text into vectors for semantic search, RAG, and similarity tasks. OpenAI-compatible and runs on European infrastructure.

The `/v1/embeddings` endpoint converts text into dense numerical vectors that capture semantic meaning. Use embeddings to power semantic search, Retrieval-Augmented Generation (RAG) pipelines, document clustering, duplicate detection, and cross-lingual similarity tasks. The endpoint is OpenAI-compatible, so any library or framework that targets the OpenAI embeddings API works out of the box by pointing `base_url` at Meliai. Embedding requests default to the `:price` routing flavor, keeping bulk vectorisation costs low.

## Endpoint

```
POST https://api.meliai.ai/v1/embeddings
```

**Authorization:** `Bearer sk-mel-<KEY>` via `Authorization` header.

***

## Parameters

<ParamField body="model" type="string" required>
  The embedding model ID to use. Check `GET /v1/models` for available embedding models. Routing flavor suffixes (e.g. `:speed`) are supported but `:price` is applied by default for embedding requests.
</ParamField>

<ParamField body="input" type="string | array" required>
  The text to embed. Pass a single string or an array of strings to embed multiple texts in one request. Batching multiple inputs in a single call is more efficient than sending them individually.
</ParamField>

<ParamField body="encoding_format" type="string">
  Format of the returned vectors:

  * `"float"` — array of 64-bit floats (default)
  * `"base64"` — base64-encoded binary representation, useful for reducing response payload size
</ParamField>

***

## Example

<CodeGroup>
  ```python Python theme={null}
  import os
  from openai import OpenAI

  client = OpenAI(
      api_key=os.environ["MELIAI_API_KEY"],
      base_url="https://api.meliai.ai/v1",
  )

  response = client.embeddings.create(
      model="<EMBEDDING_MODEL_ID>",
      input=["European AI infrastructure", "sovereign data residency"],
  )

  for item in response.data:
      print(f"Index {item.index}: vector dim {len(item.embedding)}")
  ```

  ```javascript Node.js theme={null}
  import OpenAI from "openai";

  const client = new OpenAI({
    apiKey: process.env.MELIAI_API_KEY,
    baseURL: "https://api.meliai.ai/v1",
  });

  const response = await client.embeddings.create({
    model: "<EMBEDDING_MODEL_ID>",
    input: ["European AI infrastructure", "sovereign data residency"],
  });

  for (const item of response.data) {
    console.log(`Index ${item.index}: vector dim ${item.embedding.length}`);
  }
  ```

  ```bash curl theme={null}
  curl https://api.meliai.ai/v1/embeddings \
    -H "Authorization: Bearer $MELIAI_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "<EMBEDDING_MODEL_ID>",
      "input": ["European AI infrastructure", "sovereign data residency"]
    }'
  ```
</CodeGroup>

***

## Response

```json theme={null}
{
  "object": "list",
  "model": "<EMBEDDING_MODEL_ID>",
  "data": [
    {
      "object": "embedding",
      "index": 0,
      "embedding": [0.0023, -0.0147, 0.0382, "..."]
    },
    {
      "object": "embedding",
      "index": 1,
      "embedding": [0.0051, -0.0093, 0.0271, "..."]
    }
  ],
  "usage": {
    "prompt_tokens": 8,
    "total_tokens": 8
  },
  "environment_impact": {
    "energy_kwh": 0.000008,
    "carbon_g_co2": 0.0006,
    "water_liters": 0.00003,
    "renewable_percent": 91,
    "pue": 1.2,
    "provider_id": "provider-eu-central",
    "location": "Munich, DE"
  },
  "billing_cost": {
    "energy": 0.000004,
    "credits": 0.000062,
    "paid_with": "credits"
  }
}
```

<ResponseField name="data" type="array">
  Ordered array of embedding objects, one per input string. Each contains:

  * `object` — always `"embedding"`
  * `index` — position of this item in the input array
  * `embedding` — the vector as an array of floats (or a base64 string if `encoding_format` is `"base64"`)
</ResponseField>

<ResponseField name="usage" type="object">
  Token counts for the request: `prompt_tokens` and `total_tokens`.
</ResponseField>

<ResponseField name="environment_impact" type="object">
  Per-request environmental footprint. Fields: `energy_kwh`, `carbon_g_co2`, `water_liters`, `renewable_percent`, `pue`, `provider_id`, `location`.
</ResponseField>

<ResponseField name="billing_cost" type="object">
  Itemised cost: `energy` (EUR), `credits` deducted, and `paid_with`.
</ResponseField>

***

<Tip>
  Embedding requests automatically use the `:price` routing flavor. If you need lower latency for real-time retrieval, append `:speed` to your model ID: `"<EMBEDDING_MODEL_ID>:speed"`.
</Tip>

<Note>
  All embedding computation runs on European infrastructure. Input text never leaves the EU and is never used to train models.
</Note>