Selode Mother Box - Model Comparison
Model Performance Comparison
Model Performance Comparison
Compare various AI models based on multiple performance metrics and environmental impact.
Model | Average (%) | IFEval (%) | BBH (%) | MATH (%) | GPQA (%) | MUSR (%) | MMLU-PRO (%) | CO₂ Cost (kg) |
---|---|---|---|---|---|---|---|---|
Multiverse 8B Model (LLama 3.1 based) | 30% | 81% | 31% | 17% | 8% | 11% | 32% | 0.87 |
Google 2 9B IT | 29% | 74% | 42% | 0% | 15% | 10% | 32% | 5.01 |
Qwen 7B Instruct | 27% | 76% | 35% | 0% | 5% | 8% | 37% | 2.17 |
IBM Granite 3.1 8B Instruct | 27% | 63% | 33% | 17% | 9% | 15% | 28% | 1.22 |
Mistral 8B Instruct 2410 | 22% | 59% | 26% | 7% | 5% | 11% | 25% | 0.80 |