Configure Experiment

Larger open-weight model, stronger reasoning, still served at low latency by Groq.

Open-weight flagship model with reasoning capabilities, served on Groq's LPU hardware.

Simulated GPT-4 responses for comparative analysis. Uses pre-generated outputs.

Open-source model simulation. Demonstrates open vs closed model comparison.

Always returns the exact ground-truth answer. Fixed reference point for comparison, not a model.

Provides role context and output format. Encourages organized responses.

Instructs the model to reason step-by-step before answering.