Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

GPT-5.4 Pro

OpenAI

Strongest in Reasoning (#22 of 143, on 1 of its 3 boards), weakest in Safety (#145 of 331). Above par in 4 of 4 scopes. Among the models it meets almost everywhere, it finishes behind GPT-6 Sol and Claude Opus 5.5 and ahead of GLM-5.1 and Qwen3.8 Max.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Reasoning55.2#40/189
Reasoning59.4#22/143
Safety51.3#145/331
Factual grounding54.3#41/102

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for GPT-5.4 Pro, left for the other.

Command Awon 3 · lost 0
Gemini 1.5 Prowon 3 · lost 0
GPT-3.5 Turbowon 3 · lost 0
MiniMax M3won 3 · lost 0
Llama 3.1 405B Instructwon 3 · lost 0
GLM-5.3 Flashwon 3 · lost 0
Kimi K2.7 Codewon 3 · lost 0
Magistral Smallwon 3 · lost 0
Magistral Mediumwon 3 · lost 0
Mistral Large 4won 3 · lost 0
NVIDIA Nemotron 3 Ultra 550B A55Bwon 3 · lost 0
Kimi K2.6won 3 · lost 0
GLM-5.2won 3 · lost 0
GLM-4.6won 3 · lost 0
DeepSeek V4 Flashwon 3 · lost 0
Gemini 3.6 Flashwon 3 · lost 0
Qwen3.7 Maxwon 3 · lost 0
Qwen3.5 Flashwon 3 · lost 0
Grok 4won 3 · lost 0
Qwen 3.5 Pluswon 3 · lost 0
Mistral Large 3won 4 · lost 0
O3 Prowon 4 · lost 0
DeepSeek R1won 4 · lost 0
MiniMax M2.1won 4 · lost 0
MiniMax M2.5won 4 · lost 0
Kimi K2.5won 4 · lost 0
GLM-5won 4 · lost 0
Grok 4 Fastwon 4 · lost 0
Claude Opus 4.1won 4 · lost 0
GLM-4.7won 4 · lost 0
DeepSeek V4 Prowon 4 · lost 0
Grok 4.1 Fastwon 4 · lost 0
GPT-4owon 3 · lost 1
Command R Pluswon 3 · lost 1
Llama 3.3 70B Instructwon 3 · lost 1
Llama 4 Scout Instructwon 3 · lost 1
Gemma 4 31Bwon 3 · lost 1
DeepSeek V3won 3 · lost 1
Claude Haiku 4.5won 3 · lost 1
Kimi K2won 3 · lost 1
Grok 3won 3 · lost 1
Gemini 2.5 Flashwon 3 · lost 1
Qwen3 235B A22B Instructwon 3 · lost 1
Claude Opus 4won 3 · lost 1
GPT-5 Nanowon 3 · lost 1
DeepSeek V3.1won 3 · lost 1
O4 Miniwon 3 · lost 1
Claude Sonnet 4won 3 · lost 1
GPT OSS 120Bwon 3 · lost 1
DeepSeek V3.2won 3 · lost 1
GPT-5 Miniwon 3 · lost 1
Claude Sonnet 4.5won 3 · lost 1
Gemini 2.5 Prowon 3 · lost 1
GPT-5won 3 · lost 1
GPT-5.1won 3 · lost 1
Claude Sonnet 4.6won 3 · lost 1
Claude Opus 4.5won 3 · lost 1
Gemini 3 Flashwon 3 · lost 1
Gemini 3.1 Prowon 3 · lost 1
GPT-4o Miniwon 2 · lost 1
Mistral Smallwon 2 · lost 1
GPT-4.1 Nanowon 2 · lost 1
Gemini 2.0 Flashwon 2 · lost 1
O1 Miniwon 2 · lost 1
Claude 3.5 Sonnetwon 2 · lost 1
Claude 3 Opuswon 2 · lost 1
MiniMax M2won 2 · lost 1
GPT-4 Turbowon 2 · lost 1
GPT-4.5won 2 · lost 1
GPT-4.1 Miniwon 2 · lost 1
Claude 3.7 Sonnetwon 2 · lost 1
O3 Miniwon 2 · lost 1
O1won 2 · lost 1
Mimo V2.6 Prowon 2 · lost 1
Inklingwon 2 · lost 1
Grok 3 Miniwon 2 · lost 1
Grok 4.7won 2 · lost 1
Kimi K2 Thinkingwon 2 · lost 1
GPT-6 Lunawon 2 · lost 1
O3won 2 · lost 1
GPT-5.6 Lunawon 2 · lost 1
Grok 4.5won 2 · lost 1
Gemini 3.5 Flashwon 2 · lost 1
Qwen3.8 Maxwon 2 · lost 1
GLM-5.1won 2 · lost 1
Llama 4 Maverick Instructwon 2 · lost 2
GPT-4.1won 2 · lost 2
GPT-5.4 Miniwon 2 · lost 2
GPT-5.4 Nanowon 2 · lost 2
GPT-5.2won 2 · lost 2
Claude Opus 4.6won 2 · lost 2
Claude Opus 4.7won 2 · lost 2
Gemini 3 Prowon 2 · lost 2
Gemini 4 Argonwon 2 · lost 2
GPT-6 Astrawon 2 · lost 2
Mistral Largewon 1 · lost 2
Gemini 2.5 Flash Litewon 1 · lost 2
Gemini 3.1 Flash Litewon 1 · lost 2
MiniMax M2.7won 1 · lost 2
Claude Sonnet 5won 1 · lost 2
GPT-5.6 Terrawon 1 · lost 2
Claude Opus 4.8won 1 · lost 2
GPT-5.4won 1 · lost 3
GPT-5.5won 1 · lost 3
GPT-5.6 Solwon 1 · lost 3
Grok 4.6won 0 · lost 3
Gemini 3.8 Flashwon 0 · lost 3
Claude Opus 5won 0 · lost 3
Claude Opus 5.5won 0 · lost 3
GPT-6 Solwon 0 · lost 4

§ 3 · Sources

Where the numbers come from

2 publications, 2 figures. Every one links to the page it was read from.

ARC-AGI-2 ↗1 measureread 2026-10-19
ARC-AGI-2 83.3%
Vectara Hallucination Leaderboard ↗1 measureread 2026-10-19
Vectara · Factual consistency 91.7%

Badge

PublicAI Index badge for GPT-5.4 Pro[![PublicAI Index](https://publicai.io/model-index/badge?model=gpt-5-4-pro)](https://publicai.io/model-index/m/gpt-5-4-pro)