What it does
An MCP server that runs short debates between several language models, such as GPT, Gemini, DeepSeek, Claude and local Ollama models. The models read each other's answers over several rounds, then the tool returns a short synthesis with a recommendation, the main trade-offs and the strongest disagreement. A quick mode and a multi-model code review are also included.
Use cases
- 01Compare options before choosing a technology
- 02Get a second opinion on a design question
- 03Run a multi-model review of a code change