Show HN: Have your agent consult other models (github.com)

A new tool, consult-llm, has been released, enabling users to seek second opinions from various AI models within their existing agent workflows. It supports well-known models like GPT-5.5, Gemini 3.1 Pro, and Claude Opus 4.7, among others. This tool incorporates features such as multi-turn threads, contextual file reviews, and real-time monitoring of interactions, streamlining the process of troubleshooting architectures, code changes, and complex bugs by allowing AI models to cross-check each other’s advice. Users can trigger consultations or debates through slash commands, making it easy to get diverse perspectives on challenges in development. The significance of consult-llm stems from the premise that different models, trained on varied datasets, can offer unique insights and recommendations, potentially uncovering blind spots or flawed reasoning. By facilitating inter-model consultations, this tool enhances the problem-solving capabilities of AI agents, making them more robust and collaborative. Its API and local CLI compatibility ensure that it can seamlessly integrate into existing workflows, giving developers a flexible and efficient way to leverage multiple AI resources for their tasks.