---
title: "GPT 5.4 Pro Benchmarks - RankedAGI"
description: "Detailed benchmark and metadata record for GPT 5.4 Pro."
source: "https://rankedagi.com/models/gpt-5-4-pro"
---

# GPT 5.4 Pro

| Field | Value |
| --- | --- |
| Organization | OpenAI |
| License | Proprietary |
| Type | Thinking |
| Version | Latest |
| Released | 2026-03-05 |
| Context window | 1M |
| Knowledge cutoff | 2025-08-31 |
| Input cost per million tokens | $30 |
| Output cost per million tokens | $180 |
| Model ID | gpt-5.4-pro |
| Last updated | 2026-04-29T09:58:27.831Z |

## Benchmarks

| Benchmark | Category | Value | Description | Source |
| --- | --- | --- | --- | --- |
| RankedAGI Coding | coding | 84.8% | RankedAGI Coding Score |  |
| RankedAGI Agentic | Agents | 88.7% | RankedAGI Agentic Score |  |
| BrowseComp | Agents | 89.3% | A benchmark for browsing agents | https://openai.com/index/browsecomp/ |
| RankedAGI Reasoning | reasoning | 81.2% | RankedAGI Reasoning Score |  |
| Humanity's Last Exam | reasoning | 42.7% | Multidisciplinary Reasoning (no tools) |  |
| Humanity's Last Exam w/ Tools | reasoning | 58.7% | Multidisciplinary Reasoning (with tools) |  |
| RankedAGI Math | math | 64.7% | RankedAGI Math Score |  |
| RankedAGI Overall | general | 73.4% | Overall RankedAGI score |  |
| ARC AGI 2 | reasoning | 83.3% | Abstract Reasoning Puzzles (Public) |  |
| FrontierScience Research | Research | 36.7% | Scientific Research |  |
| Finance Agent | Agents | 61.5% | Agentic Financial Analysis |  |

## Model Links

| Title | URL |
| --- | --- |
| Release | https://openai.com/index/introducing-gpt-5-4/ |