@oramacloud/secure-proxy NPM

Secure Proxy

Generating embeddings via the Secure Proxy

import { OramaProxy } from "@oramacloud/secure-proxy";

const proxy = new OramaProxy({
  api_key: "<Your Orama Secure Proxy API Key>",
});

const embeddings = await proxy.generateEmbeddings(
  "red leather shoes",
  "openai/text-embedding-ada-002"
);

console.log(embeddings);
// [-0.019633075, -0.00820422, -0.013555876, -0.011825735, 0.006641511, -0.012948156, ...]

Available models:

orama/gte-small. 384 dimensions, operated by Orama Cloud (preferred)
orama/gte-medium. 768 dimensions, operated by Orama Cloud
orama/gte-large. 1024 dimensions, operated by Orama Cloud
openai/text-embedding-ada-002. 1536 dimensions, proxied to OpenAI
openai/text-embedding-3-small. 1536 dimensions, proxied to OpenAI
openai/text-embedding-3-large. 3072 dimensions, proxied to OpenAI

Generating chat completions via the Secure Proxy

You can generate chat completions via the Secure Proxy in two different ways:

Returning a single string

import { OramaProxy } from "@oramacloud/secure-proxy";

const proxy = new OramaClient({
  api_key: "<Your Orama Secure Proxy API Key>",
});

const chatParams = {
  model: "openai/gpt-4",
  messages: [{ role: "user", content: "Who is Michael Scott?" }],
};

const response = await proxy.chat(chatParams);
console.log(response);

// "Michael Scott is a fictional character from the television show "The Office" (US version) ..."

Available models:

openai/gpt-4o
openai/gpt-4o-mini
openai/gpt-4-turbo
openai/gpt-4
openai/gpt-3.5-turbo

Returning a stream (via async iterators)

import { OramaProxy } from "@oramacloud/secure-proxy";

const proxy = new OramaClient({
  api_key: "<Your Orama Secure Proxy API Key>",
});

const chatParams = {
  model: "openai/gpt-4",
  messages: [{ role: "user", content: "Who is Michael Scott?" }],
};

for await (const message of proxy.chatStream(chatParams)) {
  console.log(message);
}

// Michael
// Scott is
// a fictional
//  character from the
//  television show
// "The
// Office" (US
// version)
// ...

Available models: