Qwen3 VL 32B Instruct
39.2%
no interval published
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,310
- Last sync
- 36 min36 minutes ago
All sources syncing normally
- Arena · Text· 6 hours ago
- Arena · WebDev· 6 hours ago
- Arena · Vision· 6 hours ago
- Arena · Agent· 6 hours ago
- GPQA Diamond· 36 minutes ago
- SWE-bench· 36 minutes ago
- Terminal-Bench· 36 minutes ago
- ARC-AGI-2· 36 minutes ago
Filter by org:
Released:
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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 200 | Qwen3 VL 32B Instruct | alibaba | Oct 2025 | $0.10 / $0.42 | 39.2% | — | — |
Showing IFBench (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.