shellChatGPT
FreeUse ChatGPT from the command-line.
FreeFree tier
Inputs: text, audio, file, urlOutputs: text, audio
About shellChatGPT
shellChatGPT is an open-source shell wrapper that provides a command-line interface for interacting with multiple AI service providers, including OpenAI's ChatGPT, Whisper, and TTS, as well as LocalAI, Ollama, Gemini, Anthropic, and others. It supports chat completions with streaming by default, text completions, insert mode for fill-in-the-middle (FIM) tasks, and audio transcription/translation via Whisper. The tool also offers optional Markdown rendering of responses, vision and multimodal model support, and the ability to process text, PDF, doc, and URL dumps. It is designed for users who prefer terminal-based workflows and need a flexible, scriptable interface to various AI models.
Key Features
Shell wrapper for multiple AI providers (OpenAI, LocalAI, Ollama, Gemini, Anthropic, Mistral, and more)
Chat completions with streaming by default
Text completions and insert mode (fill-in-the-middle) for code models
Speech-to-text (Whisper) and text-to-speech (TTS) support
Vision and multimodal model support
Optional Markdown rendering of responses
File picker and shell dump for processing text, PDF, doc, and URL content
Custom prompts and instruction prompts
Pros & Cons
Pros
- Free and open-source (MIT license)
- Supports a wide range of AI providers beyond just OpenAI
- Streaming responses for real-time interaction
- Works entirely from the command line, suitable for automation and scripting
- Includes audio input/output capabilities (Whisper, TTS)
- Active development with frequent updates
Cons
- Requires API keys from respective providers (costs may apply based on usage)
- Command-line interface may have a learning curve for non-technical users
- Dependency on external packages (e.g., jq, curl) for full functionality
- Markdown rendering and some features are optional and may need additional configuration
- Limited to text and audio modalities; no native image generation or video support
Best For
Interacting with ChatGPT and other AI models directly from the terminalAutomating AI-powered text generation in shell scriptsTranscribing audio files using Whisper from the command lineGenerating speech from text using TTS modelsProcessing and summarizing documents (PDF, doc, URLs) via AICode completion and fill-in-the-middle tasks with Mistral or other code models
FAQ
What AI providers does shellChatGPT support?
Based on the repository, shellChatGPT supports OpenAI (ChatGPT, Whisper, TTS), LocalAI, Ollama, Gemini, Anthropic, Mistral, and more. The exact list of supported providers should be verified in the project documentation.
Is shellChatGPT free to use?
The tool itself is free and open-source. However, usage of the underlying AI services (e.g., OpenAI API) may incur costs based on their pricing. Users need their own API keys.
Can I use shellChatGPT for speech recognition?
Yes, shellChatGPT supports speech-to-text via OpenAI's Whisper model, as indicated in the repository features.
Does shellChatGPT support streaming responses?
Yes, chat completions with streaming are enabled by default.
What operating systems are supported?
The tool is written in shell script and should work on Unix-like systems (Linux, macOS) with standard utilities. Windows support may require WSL or similar. This should be verified in the documentation.
How do I install shellChatGPT?
Installation instructions are provided in the GitHub repository. It typically involves cloning the repo and ensuring required packages (e.g., jq, curl) are installed. Exact steps should be checked in the README.