gpt-5.5-instant
1,491±6
1,485CI1,497
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,310
- Last sync
- 40 min40 minutes ago
All sources syncing normally
- Arena · Text· 8 hours ago
- Arena · WebDev· 8 hours ago
- Arena · Vision· 8 hours ago
- Arena · Agent· 7 hours ago
- GPQA Diamond· 40 minutes ago
- SWE-bench· 40 minutes ago
- Terminal-Bench· 40 minutes ago
- ARC-AGI-2· 40 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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 40 | gpt-5.5-instant | openai | Sep 2026 | — | 1,491 | 1,485–1,497 | #52–114 |
Showing Hard prompts exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.