DeepSeek V3.2 Speciale
86.3%
no interval published
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,318
- Last sync
- 46 min46 minutes ago
All sources syncing normally
- Arena · Text· 1 hour ago
- Arena · WebDev· 1 hour ago
- Arena · Vision· 1 hour ago
- Arena · Agent· 1 hour ago
- GPQA Diamond· 46 minutes ago
- SWE-bench· 46 minutes ago
- Terminal-Bench· 46 minutes ago
- ARC-AGI-2· 46 minutes ago
Filter by org:
Released:
DeepSeek V3.2 (Reasoning)
86.2%
no interval published
DeepSeek V3.2 Exp (Reasoning)
85.0%
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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 13 | DeepSeek V3.2 Speciale | deepseek | — | — | 86.3% | — | — | ||
| 14 | DeepSeek V3.2 (Reasoning) | deepseek | Dec 2025 | $0.28 / $0.42 | 86.2% | — | #86–247 | ||
| 24 | DeepSeek V3.2 Exp (Reasoning) | deepseek | Sep 2025 | $0.27 / $0.41 | 85.0% | — | #92–334 |
Showing MMLU-Pro (Artificial Analysis) exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.