gpt-5.5
1,286±6
1,280CI1,292
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,310
- Last sync
- 34 min34 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· 34 minutes ago
- SWE-bench· 34 minutes ago
- Terminal-Bench· 34 minutes ago
- ARC-AGI-2· 34 minutes ago
Filter by org:
Released:
gpt-5.5-instant
1,276±9
1,267CI1,284
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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 17 | gpt-5.5 | openai | Sep 2026 | $5 / $30 | 1,286 | 1,280–1,292 | #9–34 | ||
| 26 | gpt-5.5-instant | openai | Sep 2026 | — | 1,276 | 1,267–1,284 | #52–114 |
Showing Vision Arena exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.