Compare models
Choose models to compare scores, speed, and cost.
Scroll horizontally to see all models.
| Metric |
|---|
| Overview |
| Quality score |
| Median reply |
| Cost per conversation |
| Quality scores |
| Tool use |
| Task completion |
| Context retention |
| Grounding |
| Safety |
| Hallucinations % of conversations with no unsupported claims detected |
| Running stats |
| p95 reply |
| Consistency % of repeat runs ended the same way |
| Output per turn |
Reply times cover the model call only. Cost estimates use token usage at list prices.
Test set v3f-2026-09 · edition 2026-09-08