VERIFIEDOct 142026
Is GPT-5.4 Pro state of the art?
It used to be.Was SOTA in Math.
As of Oct 14, 2026, GPT-5.4 Pro from OpenAI. Was SOTA in Math. Was SOTA in Math from 2026-04-16 to 2026-09-28. Currently ranked #6.
Why
The ruling, in plain English
#6 on Humanity's Last Exam; missing GPQA Diamond and LMArena Text (style-controlled, hard prompts)
Receipts
Evidence
| Source | Score | Date | Type | Receipt |
|---|---|---|---|---|
| Humanity's Last Exam gpt-5.4-pro-2026-03-05_unknown | 0.4432 0.4–0.5 | 2026-03-05 | INDEP. | snap_fee7890b5dd0 |
| GPQA Diamond gpt-5.4-pro-2026-03-05_xhigh · xhigh | 0.946 0.9–1.0 | 2026-03-05 | INDEP. | snap_fee7890b5dd0 |
| SimpleQA Verified gpt-5.4-pro-2026-03-05_xhigh · xhigh | 0.463 0.4–0.5 | 2026-03-05 | INDEP. | snap_fee7890b5dd0 |
| ARC-AGI gpt-5.4-pro-2026-03-05_xhigh · xhigh | 0.9450000000000001 | 2026-03-05 | INDEP. | snap_fee7890b5dd0 |
Reign history
Where it sits in the succession
Math
Gemini 1.5 Pro 002
Gemini 2.5 Pro Exp 03.25
Gemini 2.5 Pro Exp 03.25
O3
O4 Mini
Gemini 2.5 Pro Exp 03.25
Gemini 2.5 Pro Preview 05.06
O3
O4 Mini
Gemini 2.5 Pro Exp 03.25
Gemini 2.5 Pro Preview 05.06
Gemini 2.5 Pro Preview 06.05
O3
O4 Mini
Gemini 2.5 Pro Preview 06.05
GPT-5
GPT-5 Mini
O3
GPT-5
GPT-5 Pro
Gemini 3 Pro Preview
GPT-5 Pro
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview
GPT-5.4 Pro
Muse Spark
GPT-6 Astra
Honest about the gaps
Where it is not the crown
| Category | This model | Why |
|---|---|---|
| Reasoning | Not enough data | #6 on Humanity's Last Exam; missing GPQA Diamond and LMArena Text (style-controlled, hard prompts) |
| Math | Was SOTA | Was SOTA in Math from 2026-04-16 to 2026-09-28. Currently ranked #6. |
| Science | #6 in this category | Last listed at #6 in Science on 2026-03-05. That board does not award a crown. |
| Factuality | Not enough data | #34 on SimpleQA Verified; missing LMArena Text factuality |
| Abstraction (ARC) | #8 in this category | Last listed at #8 in Abstraction (ARC) on 2026-03-05. That board does not award a crown. |