deepseek-v3.2-exp-thinking
1,446±11
1,436CI1,457
aiqbench · LLM benchmarks
1,446±11
1,436CI1,457
1,437±3
1,434CI1,440
1,436±6
1,430CI1,442
One grid, every column. Pick axes in the Axes panel — one axis ranks like a single board, several blend into a weighted composite, all of them read as the index. Rank is assigned before any filter and never renumbers.
| Model | Org | ± CI | Swing | Add to compare | |||||
|---|---|---|---|---|---|---|---|---|---|
| 69 | deepseek-v3.2-exp-thinking | deepseek | Oct 2026 | — | 1,446 | 1,436–1,457 | #77–181 | ||
| 90 | deepseek-v3.2 | deepseek | Dec 2025 | $0.28 / $0.42 | 1,437 | 1,434–1,440 | #86–247 | ||
| 92 | deepseek-v3.2-exp | deepseek | Sep 2025 | $0.27 / $0.41 | 1,436 | 1,430–1,442 | #92–334 | ||
| 102 | deepseek-v3.2-thinking | deepseek | Oct 2026 | — | 1,432 | 1,429–1,436 | #103–345 |
Showing Factuality Arena exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.