GPT-5.4vsGLM-5.3
GPT-5.4 | GLM-5.3 | |
|---|---|---|
| Specifications | ||
Parameters | — | 743B |
| Benchmarks | ||
Nonsense detection BullshitBench v2Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 48% | — |
Agentic coding DeepSWE 1.1Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. (Version 1.1 of the test.) Higher is better. | — | 66.9% |
Next.js coding Next.js EvalsVercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | 83% | — |
Agentic terminal coding Terminal-Bench 3.0command-line task completion (v3.0, much harder task set) | — | 28.3% |
Agentic terminal coding Terminal-Bench 2.0Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? (Version 2.0 of the test.) Higher is better. | 75.1% | — |
Software engineering Expert-SWE (Internal)OpenAI's private set of expert-level software-engineering problems. Higher is better. | 68.5% | — |
General tool use ToolathlonTests how well the AI uses everyday real-world tools and apps to get things done. Higher is better. | 54.6% | — |
Web browsing BrowseCompCan the AI browse the web and track down hard-to-find answers? Higher is better. | 82.7% | — |
Cybersecurity CyberGymTests the AI on cybersecurity challenges — finding and exploiting software weaknesses inside a safe sandbox. Higher is better. | 79% | 84.5%Best |
Cybersecurity ExploitBenchA 'capability ladder' for security research, built by CMU researchers: the AI is given known bugs in Chrome's V8 engine and scored on how far it gets toward a working exploit inside a research sandbox — from understanding the patch to triggering a crash. Higher is better. | — | 54.4% |
Cybersecurity ExploitGym · 6-hour budgetCan an AI agent turn a known software vulnerability into a working attack in a controlled lab? Built by MPI-SP researchers, the score is how many of 898 real cases (userspace programs, the V8 engine, the Linux kernel) it cracks — here with a 6-hour compute budget per case. Higher is better. | — | 130 |
Cybersecurity ExploitGym · 2-hour budgetCan an AI agent turn a known software vulnerability into a working attack in a controlled lab? Built by MPI-SP researchers, the score is how many of 898 real cases (userspace programs, the V8 engine, the Linux kernel) it cracks — here with a 2-hour compute budget per case. Higher is better. | — | 105 |
Multidisciplinary reasoning Humanity's Last Exam · with toolsHumanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | — | 62.5% |
Advanced math FrontierMath · Tier 1–3Very hard, research-level math problems. Tiers 1–3 are the (still extremely difficult) lower tiers. Higher is better. | 47.6% | — |
Advanced math FrontierMath · Tier 4Very hard, research-level math problems. Tier 4 is the hardest — close to what professional research mathematicians tackle. Higher is better. | 27.1% | — |
Science GPQA DiamondGraduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 92.8% | — |
Agentic computer use OSWorld-VerifiedCan the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Higher is better. | 75% | — |
Agentic computer use Agent's Last ExamA hard set of desktop and operating-system tasks an AI agent has to finish by looking at the screen and working the machine itself. The score is the share it passes outright — partial credit does not count. Higher is better. | — | 28.5% |
Business workflows AutomationBenchTests whether the AI can run real multi-step business workflows — the kind of end-to-end office processes companies want to automate — from start to finish. Higher is better. | — | 48.2% |
Knowledge work GDPval-AA v2economically valuable knowledge work (v2, re-based Elo) | — | 1769 |
Knowledge work GDPval (win/tie rate)How often the AI's work matches or beats a human expert's on real knowledge-work tasks. Higher is better. | 83% | — |
Community preference Arena Elo (Text)Real people chat with two anonymous AIs side by side and vote for the answer they prefer. Votes become a chess-style Elo rating on arena.ai — it measures which AI people actually like, not test scores. Higher is better. | 1476 | — |
Community preference (code) Arena Elo (Code)Like the text arena, but people vote on which AI writes better code. The votes become a chess-style Elo rating on arena.ai. Higher is better. | 1462 | — |
| Overview | ||
| Company | OpenAI | Z.ai |
| Release date | Mar 5 2026 | Aug 14 2026 |
| Access | Proprietary | Proprietary |
Which is better: GPT-5.4 or GLM-5.3?
GLM-5.3 leads GPT-5.4 on 1 of the 1 benchmark they both report (CyberGym). GPT-5.4 shipped 162 days before GLM-5.3, so benchmark comparisons should account for the intervening progress.
Published specifications for these two models are limited — see each model page for the latest details.
On CyberGym, GLM-5.3 leads at 84.5% vs GPT-5.4 at 79%.
Frequently asked questions
GPT-5.4 was released by OpenAI on Mar 5 2026.
GLM-5.3 was released by Z.ai on Aug 14 2026.