GPT-5.4 ProvsGPT-5-Pro

Across 5 shared benchmarks, GPT-5.4 Pro leads overall: GPT-5.4 Pro wins 5, GPT-5-Pro wins 0, with 0 ties and an average score difference of +26.84.

OpenAI
GPT-5.4 Pro

OpenAI · 2026-03-05 · Multimodal model

OpenAI
GPT-5-Pro

OpenAI · 2025-08-07 · Reasoning model

GPT-5.4 Pro5 wins(100%)(0%)0 winsGPT-5-Pro

Benchmark scores

Grouped by capability, sorted by largest gap within each. 5 shared benchmarks.

General Knowledge

GPT-5.4 Pro 4/4
BenchmarkGPT-5.4 ProGPT-5-ProDiff
ARC-AGI-283.306 / 59Thinking High (No Tools)1832 / 59+65.30
ARC-AGI94.505 / 65Thinking High (No Tools)70.2027 / 65+24.30
HLE58.703 / 157Thinking High (With Tools)4245 / 157+16.70
GPQA Diamond94.402 / 178Thinking High (No Tools)89.4023 / 178+5

Math and Reasoning

GPT-5.4 Pro 1/1
BenchmarkGPT-5.4 ProGPT-5-ProDiff
FrontierMath - Tier 437.505 / 80Normal (With Tools + Internet)14.6023 / 80+22.90

Specs

FieldGPT-5.4 ProGPT-5-Pro
PublisherOpenAIOpenAI
Release date2026-03-052025-08-07
Model typeMultimodal modelReasoning model
ArchitectureDenseDense
ParametersNot availableNot available
Context length1M400K
Max output125K125K

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemGPT-5.4 ProGPT-5-Pro
Text input$30 / 1M tokensNot public
Text output$180 / 1M tokensNot public

One or both models have incomplete public pricing.

Summary

  • GPT-5.4 Proleads in:General Knowledge (4/4), Math and Reasoning (1/1)

On average across the 5 shared benchmarks, GPT-5.4 Pro scores 26.84 higher.

Largest single-benchmark gap: ARC-AGI-2 — GPT-5.4 Pro 83.30 vs GPT-5-Pro 18 (+65.30).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.