Introduction to "Chat with GPT"
"Chat with GPT" is an innovative open-source application designed to enhance your interaction with ChatGPT by providing additional features and customization options. This unofficial app brings together the ChatGPT language model from OpenAI and ElevenLabs' text-to-speech capabilities, offering users a unique experience with a realistic human voice.
For those eager to explore this tool, it offers a hosted version available at: Chat with GPT. If you prefer taking control, you can also opt to self-host using Docker.
Developed using TypeScript and React, this app leverages the latest ChatGPT API from OpenAI. Community involvement is highly encouraged, and the developers welcome contributions through pull requests.
Key Features
"Chat with GPT" is packed with a range of impressive features:
- Fast Response Times: Engage in conversations with minimal delay.
- Search Functionality: Quickly locate previous chat discussions effortlessly.
- System Prompt Customization: Modify the hidden prompt that is presented to the AI before your input.
- Temperature Adjustment: Tweak the creativity and randomness of responses by changing the Temperature setting, where a higher value results in more imaginative outputs.
- Realistic Human Voice: Enhance the AI's spoken replies by integrating your ElevenLabs text-to-speech account or using your browser's in-built text-to-speech.
- Speech Recognition Technology: Utilize OpenAI Whisper for voice input.
- Public Sharing: Share cherished chat sessions through public URLs.
- Copy-and-Paste Functionality: Effortlessly duplicate ChatGPT messages.
- Message Editing: Make adjustments to your input post-submission.
- Message Regeneration: Request a new output from ChatGPT when needed.
- Comprehensive Markdown Support: Includes features for code, tables, and equations.
- Cost-effective API Usage: Pay based on consumption with the ChatGPT API.
Utilizing Your Own API Keys
OpenAI Integration
To begin interacting with "Chat with GPT," users need to input their OpenAI API key within the app’s settings. On the homepage, the prompt "Connect your OpenAI account to get started" will help guide you through this process. It’s important to note that your API key is safely stored on your device and only shared with OpenAI, with costs billed on a usage basis, separate from any existing ChatGPT subscriptions.
ElevenLabs Integration
For the realistic text-to-speech feature, users will need to incorporate their ElevenLabs API key. This can be added effortlessly by selecting "Play" adjacent to any chat message. Similar to OpenAI, the API key remains securely on your device with no external transmission apart from ElevenLabs.
Running on Your Own Computer
For users desiring full control and ownership, the app can be self-hosted using Docker with the following command:
docker run -v $(pwd)/data:/app/data -p 3000:3000 ghcr.io/cogentapps/chat-with-gpt:release
Post setup, access the app by visiting http://localhost:3000.
Storing API Keys on the Server
As a convenience, you can opt to store API keys directly on your machine. Exercise caution if others can access your environment as they could use your credentials. Store the keys in a config.yaml file located in your data directory with content like:
services:
openai:
apiKey: (your api key)
elevenlabs:
apiKey: (your api key)
After configuring, restarting the server is necessary, with login required.
Updating the Application
To keep "Chat with GPT" current with the latest features and fixes, update using:
docker pull ghcr.io/cogentapps/chat-with-gpt:release
License Information
"Chat with GPT" is available under the MIT license, allowing broad usage and modification opportunities. For complete details, please refer to the LICENSE file.