Grok 4.5vsGLM-5.3
Grok 4.5 | 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. | 54% | — |
Prompt injection robustness Gray Swan IPI · k = 1Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. Gray Swan's indirect prompt injection benchmark measures how often such an attack succeeds when the attacker gets a single try. Lower is better. | 13.4% | — |
Prompt injection robustness Gray Swan IPI · k = 10Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 10 tries and counts an attack as successful if any of them works. Lower is better. | 54.2% | — |
Prompt injection robustness Gray Swan IPI · k = 15Attackers hide malicious instructions inside content the AI reads — a web page, an email, a document — and try to hijack what it does. This variant gives the attacker 15 tries and counts an attack as successful if any of them works. Lower is better. | 60.8% | — |
Agentic coding SWE-Bench ProCan the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better. | 64.7% | — |
Multilingual coding SWE-Bench MultilingualLike SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better. | 78% | — |
Agentic coding CursorBench v3.2Cursor's own test of harder, real-world coding tasks inside a code editor, on the refreshed v3.2 task set. Scores aren't comparable with v3.1. Higher is better. | 66.7% | — |
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. | 54% | 66.9%Best |
Agentic coding DeepSWE 1.0Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. Higher is better. | 62% | — |
Agentic coding FrontierCode v1.1 (Extended) · extended splitfrontier-difficulty agentic coding tasks (v1.1, extended split) | 56.6% | — |
Expert software engineering APEX-SWEexpert-level software-engineering tasks (AI Productivity Index) | 53.6% | — |
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 computer work Frontier-Bench v0.1A hard, ever-evolving set of real computer tasks — coding, system administration, data work, and more — that an AI agent has to complete on its own. Run by the Harbor / Laude Institute team as the successor to Terminal-Bench (v0.1 is the first release of the task set). The score is the share of tasks solved. Higher is better. | 17.8% | — |
Agentic terminal coding Terminal-Bench 3.0command-line task completion (v3.0, much harder task set) | 15.7% | 28.3%Best |
Agentic terminal coding Terminal-Bench 2.1Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 83.3% | — |
Expert agentic work APEX-Agentsexpert-level agentic work tasks (AI Productivity Index) | 47.1% | — |
Cybersecurity CyberGymTests the AI on cybersecurity challenges — finding and exploiting software weaknesses inside a safe sandbox. Higher is better. | — | 84.5% |
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% |
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% |
Agentic legal work Harvey's Legal Agent BenchmarkHarvey's test of whether an AI agent can complete real legal work — drafting and reviewing documents, working with spreadsheets and presentations, and navigating files the way a lawyer's assistant would. Higher is better. | 12.92% | — |
Medical admin work MedScribeCan the AI support doctors with their administrative work, like notes and paperwork? Created by Vals AI. Higher is better. | 86.88% | — |
Overall intelligence AA Intelligence IndexArtificial Analysis composite intelligence index across evals | 56 | — |
Knowledge work GDPval-AA v2economically valuable knowledge work (v2, re-based Elo) | 1526 | 1769Best |
Knowledge work AA-BriefcaseArtificial Analysis agentic office-work eval (Elo) | 1313 | — |
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. | 1468 | — |
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. | 1549 | — |
| Overview | ||
| Company | SpaceXAI | Z.ai |
| Release date | Jul 8 2026 | Aug 14 2026 |
| Access | Proprietary | Proprietary |
Which is better: Grok 4.5 or GLM-5.3?
GLM-5.3 leads Grok 4.5 on 3 of the 3 benchmarks they both report (DeepSWE 1.1, Terminal-Bench 3.0, GDPval-AA v2). Grok 4.5 shipped 37 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 DeepSWE 1.1, GLM-5.3 leads at 66.9% vs Grok 4.5 at 54%. On Terminal-Bench 3.0, GLM-5.3 leads at 28.3% vs Grok 4.5 at 15.7%. On GDPval-AA v2, GLM-5.3 leads at 1769 vs Grok 4.5 at 1526.
Frequently asked questions
Grok 4.5 was released by SpaceXAI on Jul 8 2026.
GLM-5.3 was released by Z.ai on Aug 14 2026.
GLM-5.3 leads on DeepSWE 1.1 — Grok 4.5 54% vs GLM-5.3 66.9%.