Development

A model matches human professionals 83% of the time

development2026-03-06

Claim

OpenAI's GPT-5.4 Thinking and Pro arrived with native computer use, a million-token context and an 83.0% GDPval score, matching or exceeding human professionals at knowledge work 83% of the time, alongside 57.7% on SWE-Bench Pro and 67.3% on WebArena-Verified.

Reported metric: 83.0% GDPval

Reporting date: 2026-03-06 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to March 6, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

About

GPT-5.4 Pro

Reported in

Welcome to March 6, 2026

Actor

OpenAI

Evidence for

Supersedes

Parallelizing drafting and verification gives 5x faster inference

Referenced by

Welcome to March 6, 2026

Cite or correct this record

Citation formatCitation text

A model matches human professionals 83% of the time
Innermost Loop KB — secondary record
Reporting date: 2026-03-06
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-03-06-gdpval-83-percent
Claim: OpenAI's GPT-5.4 Thinking and Pro arrived with native computer use, a million-token context and an 83.0% GDPval score, matching or exceeding human professionals at knowledge work 83% of the time, alongside 57.7% on SWE-Bench Pro and 67.3% on WebArena-Verified.
Attributed sources:
- Welcome to March 6, 2026 — https://theinnermostloop.substack.com/p/welcome-to-march-6-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.