Deterministic Lexical Feature Embeddings
Create reproducible normalized 256-dimensional lexical feature-hash vectors. These are not neural semantic embeddings.
Preview request
The preview path is rate-limited and does not charge. Production requests use the same parameters and return a standard x402 challenge when no payment signature is present.
curl --request POST \
"https://ai-data-marketplace-1042299154756.us-central1.run.app/api/v1/embed?preview=true" \
--header "content-type: application/json" \
--data '{"texts":["Blue running shoes","Red sports car"],"model":"text-embedding-3-small"}'Input contract
| Parameter | Type | Required | Description |
|---|---|---|---|
texts | array | Yes | Array of strings to embed. |
model | string | No | Embedding model to use. |
Guaranteed response metadata
Every successful product response identifies how the result was produced. Product-specific output fields are documented in the paid OpenAPI contract and can be inspected through the preview request where available.
{
"marketplace_metadata": {
"data_mode": "computed_service",
"billable": true,
"availability": "operational",
"source": "deterministic feature-hash vectorizer"
}
}Agent payment flow
Request the product, read the base64 PAYMENT-REQUIRED challenge, sign the exact Base-USDC authorization with an x402 buyer client, and retry using PAYMENT-SIGNATURE. A successful response includes PAYMENT-RESPONSE.