Repository navigation
Local Model Bench 0.1.0
Initial public release of Local Model Bench.
Highlights:
- Graphical benchmark UI for LM Studio models.
- Objective 200-case frontier-style benchmark suite.
- Balanced category and per-test comparison views.
- Model metadata display for quantization, size, architecture, and speed.
- Reasoning capability detection and requested maximum reasoning level.
- Batch mode for testing all not-yet-tested loaded model variants.
- Safe batch model switching with explicit LM Studio unload confirmation.
- Windows and Ubuntu/Linux starter scripts.