$ loading_
Compare multiple LLM answers blindly and score them before revealing models.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "llm-council" yet — see the docs or source repo.
Send this question to multiple configured models and hide their names while returning independent answers: 'Compare when to use RAG versus fine-tuning for enterprise knowledge-base QA.' Then rank the answers with blind scoring and reveal the models afterward.
Anonymous answers, blind evaluation results, and the final mapping back to each model.
Ask multiple models to explain the potential bug and fix for this code, but do not show model identities first; I want to score the answers for clarity and correctness before seeing which model produced each one.
Multiple code analysis responses for anonymous comparison, followed by model reveal after scoring.
Start a direct one-to-one chat with one connected model and continue discussing 'how to design an evaluation workflow for a local MCP tool' while keeping context.
A continuous single-model conversation for deeper follow-up and iteration.
Developers or researchers can collect independent answers to the same question, review them anonymously, and reveal the models afterward to reduce brand bias. This is useful for model selection and response quality comparison.
Product managers or teams can send the same task to multiple model CLIs and use blind scoring to judge which output best fits their needs. The workflow suits quick comparative testing in a local MCP setup.
When multi-model comparison is not needed, users can chat directly one-to-one with a specific model and continue asking about the same topic. This is useful for deeper exploration after an evaluation round.
It is a local MCP server for running anonymous multi-model councils. You can ask the same question to multiple model CLIs, review independent answers blindly, and reveal model identities afterward.
Based on the description, it depends on a local MCP server environment and works with multiple model CLIs. For exact installation and configuration steps, see the source repository.
No. The description also mentions direct one-to-one chat, so besides blind multi-model evaluation, it can also be used to talk with a single model directly.
Run local multi-model deliberation and synthesis for better coding decisions on Mac.
Ask other models for opinions and compare their answers quickly.
Run local multi-model deliberation and synthesis on Mac for AI coding workflows.
Connect Claude Code to external LLMs for consultation, verification, and web research.
Get real-time AI model pricing, capabilities, and context window details.
Compare AI models across providers by cost, performance, and capabilities.