API Reference

Hush AI API Documentation

OpenAI-compatible chat completions API. Your data never leaves hardware we own.

Authentication Chat Completions Streaming Models Errors

All API requests are processed in memory on hardware we own. Nothing is stored, nothing is logged, nothing is trained on. This applies to every endpoint and every model.

Authentication

All API requests require an API key passed in the Authorization header as a Bearer token.

Authorization: Bearer YOUR_API_KEY

Get your API key by creating an account. API keys are available on all plans.

Keep your key secret. Do not expose it in client-side code, public repositories, or browser requests. Use it only in server-side code or secure environments.

Base URL

https://hush-ai.uk/v1

All endpoints are prefixed with this base URL. The API accepts application/json requests and returns application/json responses (or text/event-stream for streaming).

Endpoints

Create chat completion

POST /v1/chat/completions

Creates a chat completion for the given messages. This is the primary endpoint and follows the OpenAI chat completions format exactly.

Request body

Parameter	Type		Description
model	string	Required	Model ID to use: `omega`, `prism`, or `lambda`.
messages	array	Required	Array of message objects. Each has `role` (`system`, `user`, or `assistant`) and `content` (string).
temperature	number	Optional	Sampling temperature, 0 to 2. Higher values make output more random. Default: 1.
max_tokens	integer	Optional	Maximum number of tokens to generate in the response.
top_p	number	Optional	Nucleus sampling. Consider tokens with top_p probability mass. Default: 1.
stream	boolean	Optional	If `true`, partial message deltas are sent as server-sent events. Default: `false`.
stop	string or array	Optional	Up to 4 sequences where the API will stop generating.
frequency_penalty	number	Optional	Penalise tokens based on frequency in the text so far, −2.0 to 2.0. Default: 0.
presence_penalty	number	Optional	Penalise tokens based on whether they appear in the text so far, −2.0 to 2.0. Default: 0.

Example request

curl https://hush-ai.uk/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "omega",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Explain GDPR Article 28 in one paragraph."}
    ],
    "temperature": 0.7,
    "max_tokens": 256
  }'

Response

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1722100000,
  "model": "omega",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "GDPR Article 28 governs the relationship between..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 28,
    "completion_tokens": 95,
    "total_tokens": 123
  }
}

Response fields

Field	Type	Description
id	string	Unique identifier for this completion.
object	string	Always `chat.completion`.
created	integer	Unix timestamp of when the completion was created.
model	string	The model used for this completion.
choices	array	Array of completion choices. Each contains `index`, `message` (with `role` and `content`), and `finish_reason`.
usage	object	Token usage: `prompt_tokens`, `completion_tokens`, `total_tokens`.

Streaming

Streaming responses

Set "stream": true in your request to receive partial responses as server-sent events (SSE). The response uses Content-Type: text/event-stream.

Example streaming request

curl https://hush-ai.uk/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -N \
  -d '{
    "model": "omega",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": true
  }'

Stream format

Each event is a JSON object prefixed with data: . The stream ends with data: [DONE].

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1722100000,"model":"omega","choices":[{"index":0,"delta":{"content":"Hello"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1722100000,"model":"omega","choices":[{"index":0,"delta":{"content":"!"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1722100000,"model":"omega","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

Handling streams in Python

from openai import OpenAI

client = OpenAI(
    base_url="https://hush-ai.uk/v1",
    api_key="YOUR_API_KEY"
)

stream = client.chat.completions.create(
    model="omega",
    messages=[{"role": "user", "content": "Hello"}],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

Handling streams in Node.js

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://hush-ai.uk/v1",
  apiKey: "YOUR_API_KEY",
});

const stream = await client.chat.completions.create({
  model: "omega",
  messages: [{ role: "user", content: "Hello" }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || "");
}

Models Endpoint

List models

GET /v1/models

Returns the list of available models.

Example request

curl https://hush-ai.uk/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Response

{
  "object": "list",
  "data": [
    {"id": "omega", "object": "model", "owned_by": "hush-ai"},
    {"id": "prism", "object": "model", "owned_by": "hush-ai"},
    {"id": "lambda", "object": "model", "owned_by": "hush-ai"}
  ]
}

For full model specifications, see the models page.

Error Handling

Error codes

The API returns standard HTTP status codes. Error responses include a JSON body with error.message and error.type.

{
  "error": {
    "message": "Invalid API key provided.",
    "type": "authentication_error",
    "code": "invalid_api_key"
  }
}

Code	Meaning	What to do
400	Bad request: malformed JSON or missing required fields.	Check your request body. Ensure `model` and `messages` are present and correctly formatted.
401	Unauthorised: missing or invalid API key.	Check the `Authorization` header. Ensure the key starts with the correct prefix and has not been revoked.
403	Forbidden: the key does not have permission for this resource.	Contact support.
404	Not found: the endpoint or model does not exist.	Check the URL and model ID. Valid models: `omega`, `prism`, `lambda`.
429	Rate limited: too many requests.	Back off and retry with exponential backoff. Check the `Retry-After` header if present.
500	Internal server error.	Retry after a brief pause. If persistent, contact support.
503	Service unavailable: the API is temporarily overloaded or in maintenance.	Retry with backoff. Check the changelog for scheduled maintenance.

Request Format

Content type

All requests must use Content-Type: application/json with UTF-8 encoding. Responses are application/json for non-streaming requests and text/event-stream for streaming requests.

Developer overview · Models · API pricing · Quickstart · Changelog

Security · Verify us · Hard questions, answered straight

Hush AI API Documentation

Authentication

Base URL

Create chat completion

Request body

Example request

Response

Response fields

Streaming responses

Example streaming request

Stream format

Handling streams in Python

Handling streams in Node.js

List models

Example request

Response

Error codes

Content type

Related pages