llm-panel is an MIT-licensed Python package and CLI for sending one question to several language models independently. It presents complete responses, parallel judge assessments, failure classifications, cost information, and an anonymized rebuttal round, with support for Ollama and OpenRouter.
In the LLM evaluation & benchmarks space, llm-panel takes a focused approach. It focuses on comparing and evaluating answers from multiple language models consistently in one workflow. It is built as an open-source project for developers and AI evaluators. llm-panel is open source under the MIT license. llm-panel is available on the command line, and it can be self-hosted.
llm-panel first shipped in 2026. Among its 7 catalogued features are multi-model queries, parallel judging, and failure classes.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do