gllm

The main gllm module. This module contains high-level functions and some types for interacting with OpenAI compatible APIs.

It includes functionality for creating chat completion requests, handling responses, and managing API authentication. The client supports customizable base URLs and temperature parameters for controlling response randomness.

Example

import gllm
import gllm/types/message
import glenvy/dotenv
import glenvy/env
import gllm

let assert Ok(api_key) = env.string("OPENROUTER_API_KEY")
let base_url = "https://openrouter.ai/api/v1"

let client = gllm.Client(api_key, base_url)

let messages = [
  gllm.new_message("system", "You are a helpful assistant."),
  gllm.new_message("user", "Hello, how are you?")
]

gllm.completion(client, "openai/gpt-oss:20b", messages, 0.7)

Types

Client

</>

Represents a client for making API requests to OpenAI-compatible endpoints.

Parameters

api_key: API KEY from an OpenAI compatible API
base_url: URL of OpenAI compatible API (example: “https://openrouter.ai/api/v1”)

pub type Client {
  Client(api_key: String, base_url: String)
}

Constructors

Client(api_key: String, base_url: String)

Values

completion

</>

pub fn completion(
  client client: Client,
  model model: String,
  messages messages: List(message.Message),
  temperature temperature: Float,
) -> Result(chat_completion.ChatCompletion, api_error.ApiError)

Sends a chat completion request to the API.

Parameters

client: The API client containing authentication and endpoint information
model: The model identifier to use for completion (e.g., “gpt-3.5-turbo”)
messages: A list of messages representing the conversation history
temperature: Controls randomness in the response (0.0 to 2.0, lower = more focused)

Returns

Returns Ok(ChatCompletion) on success, or Error(ApiError) on failure.

Errors

ApiError: Unexpected JSON response from the API or Http error

Example

let client = Client("api-key", "https://api.openai.com")
let messages = [new_message("user", "Hello")]

completion(client, "gpt-3.5-turbo", messages, 0.7)
// -> Ok(ChatCompletion(...))

message_to_json

</>

pub fn message_to_json(message: message.Message) -> json.Json

Converts a Message type to a JSON object for API requests.

Parameters

message: The Message to convert

Returns

Returns a JSON object containing the role and content fields.

Examples

let msg = new_message("user", "Hello")
message_to_json(msg)
// -> json.object([#("role", json.string("user")), #("content", json.string("Hello"))])

new_message

</>

pub fn new_message(
  role role: String,
  content content: String,
) -> message.Message

Creates a new message with the given role and content.

Parameters

role: The role of the message sender (e.g., “system”, “user”, “assistant”)
content: The text content of the message

Returns

Returns a Message type with optional fields set to None.

Examples

new_message("user", "Hello, world!")
// -> Message("user", "Hello, world!", None, None, None)