@codeworksh/aikit weekly downloads statistics

@codeworksh/aikit

AiKit is a TypeScript SDK that provides a unified API for working with multiple LLM providers, automatic model discovery, provider configurations, token and cost tracking, and mid-session hand-off to other models. It gives you the basic primitives for streaming LLM responses without the extra bloat, letting you handle the orchestration yourself.

Table of Contents

• Installation • Supported Providers • Quick Start • Tools • Providers and Models • CLI Tools • Model Generation • OAuth Providers • Contribute

Installation

Node >=22.0.0 is required.

Supported Providers

Built on top of the Vercel AI SDK, @codeworksh/aikit supports a wide array of providers natively:

• Anthropic (@ai-sdk/anthropic) • OpenAI (@ai-sdk/openai) • Google (@ai-sdk/google) • Google Vertex AI (@ai-sdk/google-vertex) • xAI (@ai-sdk/xai) • OpenRouter (@openrouter/ai-sdk-provider) • OpenAI Compatible APIs (@ai-sdk/openai-compatible)

Quick Start

Tools

Tools enable LLMs to interact with external systems. @codeworksh/aikit provides native support for function calling and schema validation using TypeBox.

Providers and Models

@codeworksh/aikit uses a registry to fetch model specifications and metadata.

CLI Tools

@codeworksh/aikit ships with a CLI tool for managing local metadata and authentication.

You can invoke it using npx aikit or pnpm aikit.

Model Generation

Generate a models.gen.json file fetching the latest metadata about providers and their available models from models.dev.

OAuth Providers

Manage OAuth credentials for providers that require it, such as OpenAI Codex and GitHub Copilot.

Contribute

@codeworksh/aikit is open to community contribution. Please ensure you submit an issue before submitting a pull request. The project prefers open community discussion before accepting new features.

Description from the npm listing.