Using Google Gemini Chat Completion

Google’s Gemini models are available through the same OpenAI-compatible request shape as the other chat completion endpoints, so switching between providers only requires changing the URL and model name.

Personalities and built-in functions are features of the OpenAI endpoint and are not available here.

Request

URL: https://ai.purlabs.xyz/google/chat/completions

Request Method: POST

TS Interface


interface Request {
	model: string;
	messages: {
		role: 'system' | 'user' | 'assistant';
		content: string;
	}[];
	temperature?: number;
	top_p?: number;
	top_k?: number;
	max_tokens?: number;
	stop?: string | string[] | null;
};

Request API Reference

Parameter	Type (check TS Interface)	Required	Description
`model`	`string`	Required	The model to use, such as `gemini-pro`. Refer to the Listing Available Models page for more information.
`messages`	`object[]`	Required	The list of messages in the conversation so far.
`messages.role`	`string`	Required	The author of the message: `system`, `user`, or `assistant`.
`messages.content`	`string`	Required	The text content of the message.
`temperature`	`number`	Optional	The sampling temperature, between 0 and 1. Higher values make the output more random; lower values make it more focused and deterministic.
`top_p`	`number`	Optional	Nucleus sampling: the model considers tokens comprising the top `top_p` probability mass.
`top_k`	`number`	Optional	The model considers only the `top_k` most likely tokens at each step.
`max_tokens`	`number`	Optional	The maximum number of tokens to generate in the response.
`stop`	`string \| string[] \| null`	Optional	Up to four sequences where the model will stop generating further tokens.

Example JSON Request Body


{
    "model": "gemini-pro",
    "messages": [
        {
            "role": "user",
            "content": "Summarize the benefits of unit testing in two sentences."
        }
    ]
}

Response

TS Response Interface


interface Response {
	id: string;
	object: 'chat.completion';
	created: number;
	model: string;
	choices: [
		{
			index: number;
			message: {
				role: 'assistant',
				content: string;
			};
			finish_reason: any;
		}
	];
	overwritten: boolean;
	provider: string;
};

Example JSON Response


{
    "id": "f81d4fae7dec11d0a76500a0c91e6bf6",
    "object": "chat.completion",
    "created": 1703031002,
    "model": "gemini-pro",
    "choices": [
        {
            "index": 0,
            "message": {
                "role": "assistant",
                "content": "Unit testing catches defects early by verifying that each piece of code behaves as expected in isolation, which makes bugs cheaper and faster to fix. It also serves as living documentation and gives developers the confidence to refactor code without breaking existing functionality."
            },
            "finish_reason": "stop"
        }
    ],
    "provider": "provider-1",
    "overwritten": false
}