gpt-5.4-pro-2026-03-05
44.3%±2.0%
42.4%CI46.3%
aiqbench · LLM benchmarks
Benchmarks
- Sources
- 8
- Entries tracked
- 22,318
- Last sync
- 29 min29 minutes ago
All sources syncing normally
- Arena · Text· 11 hours ago
- Arena · WebDev· 11 hours ago
- Arena · Vision· 11 hours ago
- Arena · Agent· 11 hours ago
- GPQA Diamond· 29 minutes ago
- SWE-bench· 29 minutes ago
- Terminal-Bench· 29 minutes ago
- ARC-AGI-2· 29 minutes ago
Filter by org:
Released:
gpt-5.4-2026-03-05
36.2%±1.9%
34.4%CI38.1%
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 | |||||
|---|---|---|---|---|---|---|---|---|---|
| 6 | gpt-5.4-pro-2026-03-05 | OpenAI | Mar 2026 | $30 / $180 | 44.3% | 42.4%–46.3% | #20–171 | ||
| 9 | gpt-5.4-2026-03-05 | OpenAI | Sep 2026 | $2.50 / $15 | 36.2% | 34.4%–38.1% | #12–27 |
Showing Humanity's Last Exam exactly as the source published it — its own score and confidence interval, no blending. Select a second axis to build a weighted composite.