DeepSeek-V3.2-Exp
94.7%
no interval published
aiqbench · LLM benchmarks
94.7%
no interval published
94.5%
no interval published
93.9%
no interval published
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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 15 | DeepSeek-V3.2-Exp | deepseek-ai | Sep 2025 | $0.27 / $0.41 | 94.7% | — | #92–334 | ||
| 18 | DeepSeek-V3.1 | deepseek-ai | Aug 2025 | — | 94.5% | — | #101–158 | ||
| 25 | DeepSeek-V3 | deepseek-ai | Dec 2024 | — | 93.9% | — | #178–644 | ||
| 27 | DeepSeek-V3.2 | deepseek-ai | Dec 2025 | $0.28 / $0.42 | 93.7% | — | #86–247 |
Showing HHEM Hallucination (Vectara) exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.