← Leaderboard
Anthropic: Claude Sonnet 4.6
anthropic/claude-sonnet-4.6 · anthropic · context 1 000 000 · in $3.00/1M · out $15.00/1M
Global Index
683
95% CI [625–741] · index v0.1.0
Per-domain scores
| Domain | Score (95% CI) | Accuracy (IRT) | Consistency | Calibration | Contam. Δ | p50 | $/1k | |
|---|---|---|---|---|---|---|---|---|
| code | 738 [584–893] | 0.527 | 1.00 | 0.99 | 0.000 | 1.2s | $6.63 | |
| instruction following | 690 [544–836] | 0.495 | 1.00 | 0.98 | 0.038 | 1.6s | $1.17 | |
| knowledge | 723 [570–877] | 0.500 | 1.00 | 0.99 | 0.000 | 1.2s | $0.518 | |
| math | 534 [446–622] | 0.310 | 1.00 | 0.71 | 0.058 | 1.1s | $1.65 | |
| multilingual | 726 [573–880] | 0.505 | 1.00 | 0.99 | 0.000 | 1.1s | $0.801 | |
| reasoning | 689 [543–835] | 0.495 | 0.98 | 0.99 | 0.038 | 1.1s | $1.77 |
Run history
- 2026-08-05v0.1.0index_fit683
- 2026-08-05v0.1.0index_fit684
- 2026-08-05v0.1.0index_fit696
- 2026-08-05v0.1.0demo_seed655