Embeddings

OpenAI-compatible embeddings under the same wallet — semantic search, RAG, and clustering with in-band cost.

POST /v1/embeddings turns text into vectors for semantic search, RAG, recommendations, deduplication, and clustering. It is wire-identical to the OpenAI Embeddings API, so the SDK works with just the base URL swap — and as everywhere on Gopuram, the exact billed cost rides in the response.

from openai import OpenAI

client = OpenAI(base_url="https://api.gopuram.net/v1", api_key=GOPURAM_API_KEY)

result = client.embeddings.create(
    model="alibaba/qwen3-embedding-0.6b",
    input=["the gopuram rises over the temple", "an unrelated sentence"],
)
print(len(result.data), "vectors")
print(result.usage)  # prompt_tokens + cost, in-band

curl https://api.gopuram.net/v1/embeddings \
  -H "Authorization: Bearer $GOPURAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "alibaba/qwen3-embedding-0.6b", "input": "the gopuram rises"}'

Request

Field	Type	Notes
`model`	string	any embedding model from the catalog
`input`	string or string[]	up to 2,048 items / 1M characters total
`encoding_format`	`"float"` \| `"base64"`	optional, passed through
`dimensions`	int	optional — models that support Matryoshka truncation honor it

Response

The OpenAI shape, plus usage.cost:

{
  "object": "list",
  "data": [{ "object": "embedding", "index": 0, "embedding": [0.0123, …] }],
  "model": "alibaba/qwen3-embedding-0.6b",
  "usage": { "prompt_tokens": 8, "total_tokens": 8, "cost": 0.000000168 }
}

Embeddings bill on input tokens only, at the rate shown per model in the catalog (margin included; the figure matches usage.cost exactly). Filter the live catalog for embedding models:

curl -s https://api.gopuram.net/v1/models \
  | jq '.data[] | select(.architecture.output_modalities == ["embedding"]) | {id, pricing}'

Cheap bulk pick: alibaba/qwen3-embedding-0.6b ($0.01/M). Highest quality: google/gemini-embedding-001 or cohere/embed-v4.0. Sending a chat model here (or an embedding model to /v1/chat/completions) returns a 400 naming the right endpoint.