This repository provides a guide and associated shortcuts for integrating ChatGPT with Siri, enabling voice-based interaction with AI models. It targets Apple users seeking to leverage ChatGPT's capabilities through their devices' native voice assistant. The primary benefit is hands-free access to advanced AI for tasks like answering questions, generating text, and even creating images.

How It Works

The project leverages Apple's Shortcuts app to create custom voice commands that interface with a third-party intermediary API. This API acts as a bridge, forwarding user queries from Siri to OpenAI's ChatGPT models (including GPT-3.5 and GPT-4) and then relaying the responses back. This approach allows for flexible integration across various ChatGPT-compatible applications beyond just Siri.

Quick Start & Requirements

• Install: Import provided .shortcut files into the Apple Shortcuts app.
• Prerequisites: An OpenAI API key (starting with "ak-") is required. A "global scientific" internet connection is recommended for faster responses, especially for users in China.
• Setup: Import shortcuts and input your API key into the shortcut's text field. Initial setup involves granting permissions.
• Links: Download shortcuts (various versions available).

Highlighted Details

• Supports both conversational AI and AI image generation (via OpenAI's official API).
• Offers multiple versions of shortcuts, including single/continuous Q&A and image generation capabilities.
• Addresses specific bugs like Siri pop-up character limits and text pop-up issues.
• The project is marked as "final" with an intermediary API now available.

Maintenance & Community

The project creator has indicated this is the final version and will no longer be updated. They encourage users to support them via "瑞幸" (a coffee brand). There are no links to community channels or roadmaps provided.

Licensing & Compatibility

The project explicitly states that the shortcuts are "开源啦" (open-sourced). However, the specific open-source license is not mentioned. The project relies on OpenAI's API, which has its own terms of service.

Limitations & Caveats

The project is marked as "final" with no further updates planned. Some users may experience slow response times without a "global scientific" connection. Certain features, like continuous conversation for image generation, may still have bugs. The project relies on a third-party intermediary API, the stability and longevity of which are not guaranteed.