gllm: A Feature-Rich LLM Command-Line Interaction Tool

gllm is a cross-platform command-line tool written in Go, designed to provide developers with a unified LLM interaction interface. It addresses the problem of API fragmentation among different LLM vendors, supporting multiple models (OpenAI, Claude, Gemini, etc.), multi-agent workflows, MCP protocol, file processing, search and retrieval, and other rich features, covering needs from simple Q&A to complex automated workflows.

With the rapid development of the LLM ecosystem, developers need to frequently interact with models from different vendors (such as OpenAI, Anthropic, Google), but the fragmentation of APIs and interaction methods across platforms creates work friction. gllm emerged as a cross-platform CLI tool that provides a unified, customizable interface, supporting scenarios like quick queries, multi-turn conversations, file processing, multi-agent workflows, etc., aiming to meet LLM interaction needs in one stop.

The core advantages of gllm include:

1. Multi-model and API compatibility: Natively supports OpenAI, Anthropic, Google Gemini APIs, allowing flexible model switching without memorizing differences;
2. Interactive REPL session: Run gllm to enter session mode, supporting context compression to avoid token limits;
3. File and attachment processing: Reference local files/directories, images, or URLs via the @ syntax, which are automatically included in context analysis.

gllm also has advanced features:

1. Search and web retrieval: The -s parameter triggers web search, answering based on the latest results and supporting URL content extraction;
2. Agent mode and multi-agent workflows: Command agents can execute system commands (Plan/Yolo modes), supporting multi-agent parallel tasks, cross-LLM handover, and context transfer;
3. MCP protocol support: Connect to external MCP servers to expand tool and data source integration capabilities;
4. Other practical features: Editor integration, prompt templates, token tracking, memory management, theme customization, API local proxy, etc.

Installation Methods:

• macOS (Homebrew): brew tap activebook/gllm && brew install gllm --cask
• Windows (Scoop): scoop bucket add gllm https://github.com/activebook/scoop-gllm.git && scoop install gllm
• Linux: curl -fsSL https://raw.githubusercontent.com/activebook/gllm/main/build/install.sh | sh It is recommended to run gllm init to configure API keys and other settings for the first use. Typical Use Cases:
• Quick query: gllm "Explain Go interfaces"
• Specify model: gllm "London attractions recommendation" -m gpt4o
• File analysis: gllm "Improve project structure: @service/ @cmd/"
• Search for latest information: gllm "Who is the current US President?" -s

gllm provides an official VS Code extension called gllm Companion, which deeply integrates the CLI tool with the IDE. It supports native inline difference display of AI code suggestion changes and can directly access workspace files, providing a convenient experience for IDE users.

gllm is designed around actual workflows, providing a unified entry point from simple Q&A to complex multi-agent collaboration, and from single-turn queries to cross-session memory. It is suitable for developers who need to switch between different models, process file attachments, and build automated workflows. Its cross-platform support and automatic update mechanism reduce long-term usage costs, making it one of the mature directions for LLM tooling.