Real head-to-head · same prompt, one shot
Claude Opus 5 vs GLM-5.2
The new Anthropic flagship — benched on all 45 one-shot builds the day it landed. vs The never-forgets agent — 1M context, open weights.
Head-to-head verdict: Claude Opus 5 wins 36–10 with 1 tie.
What I tested — same prompt, two models
I run the same fixed prompt set through every new model the day it drops — same string, one shot, single HTML file out — and I score the result 0–10 on whether it ran, how close it hit the brief, and how good it looked. Below is what came out when I gave the exact same prompts to Claude Opus 5 and GLM-5.2, side by side, on 47 shared tasks inside the Agent Operating System.
Both models were given identical prompts inside the Agent Operating System — no help, no iteration, no "best of N" tricks. I run each prompt once, save the HTML file the model produces, and score it 0–10 on whether it ran, how close it hit the brief, and how good it looked. The scoring is mine. The verdicts below are pulled from my source comparison guides at agentos.guide where I publish every score and the reasoning behind it.
Claude Opus 5 · Benched on all 45 GoldieBench tasks via API on release day, incremental deploys as scores landed.
GLM-5.2 · Default model inside Agent OS for any task that touches a long context — codebase Q&A, multi-file refactors, agent memory replay.
Side-by-side on 50 shared tasks
Click any cell to play that model's actual one-shot attempt. Medals are derived from my 0–10 scores per task (highest = 🥇, second = 🥈, third = 🥉).
Task ↓
Claude Opus 5
GLM-5.2
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Game
Page
Where Claude Opus 5 beat GLM-5.2
The tasks where I gave Claude Opus 5 a higher 0–10 score on the same prompt — with the actual commentary from my source guides.
Fireworks
Visual
Claude Opus 5 9.0
·
GLM-5.2 7.0
(+2.0)
· harbour firework spectacular
What I saw: Stunning multi-type bursts (rings, peonies, chrysanthemums, willows) over a beautifully lit skyline with reflections, moon, and stars; rich UI with auto/finale/sound controls and live stats make it a clear task winner. Only minor concern is the 22fps under heavy load, but visuall…
Aurora
Visual
Claude Opus 5 8.6
·
GLM-5.2 7.0
(+1.6)
· cinematic aurora scene
What I saw: Gorgeous WebGL curtains with realistic vertical streaks, layered mountains, starfield and water reflection make this a full cinematic scene; polished palette switcher and typography plus reflection depth push it past the field's best.
Orbit
Sim
Claude Opus 5 9.0
·
GLM-5.2 7.5
(+1.5)
· 3D N-body simulation
What I saw: Gorgeous 3D N-body sim with a glowing central star, 167 orbiting bodies with velocity trails, merges counter (12), tilted disks and companion cluster — genuine physics with softening, substeps, and merging, plus a polished camera/UI. Strong: visible dynamic simulation and beautif…
Cloth
Sim
Claude Opus 5 8.4
·
GLM-5.2 7.0
(+1.4)
· draped cloth folds
What I saw: Strong Verlet cloth showing convincing draping folds over the sphere with a colorful checkered texture, shadows, and rich controls (shapes, wind, pin, wireframe); the drape reads a touch crumpled/asymmetric rather than a clean symmetric shroud, keeping it just shy of the field's best.
Terrain
Visual
Claude Opus 5 8.3
·
GLM-5.2 7.0
(+1.3)
What I saw: Strong island terrain with convincing height-based color ramp, sandy shores, snow-capped ridges, translucent water and a rich HUD/control panel; weak point is the flat black tree cones that read as silhouettes rather than lit foliage, slightly cheapening an otherwise polished, sh…
Where GLM-5.2 beat Claude Opus 5
The tasks where I gave GLM-5.2 a higher 0–10 score on the same prompt — with the actual commentary from my source guides.
Neoncity
Game
GLM-5.2 9.0
·
Claude Opus 5 7.4
(+1.6)
· winner · cinematic
What I saw: GLM's is the most cinematic — neon towers, a setting sun, Japanese signage and a flight HUD, like a frame from a film. Opus's is a clean canyon of lit skyscrapers racing to a vanishing point. Kimi leaned into the synthwave sun and grid more than the city itself. GLM wins the skyline.
Arcade
Game
GLM-5.2 8.0
·
Claude Opus 5 6.5
(+1.5)
What I saw: All three shipped a genuinely juicy game. Opus's breakout had the most game-feel — particle bursts and a live combo. Kimi's breakout was clean and solid. GLM went its own way with fullscreen neon asteroids. The closest of the practical five.
Pool
Game
GLM-5.2 7.5
·
Claude Opus 5 6.0
(+1.5)
What I saw: 46KB · plays clean · plain
Doom
Game
GLM-5.2 8.0
·
Claude Opus 5 6.8
(+1.2)
What I saw: All three are real, playable shooters. Opus drops you in a corridor with an imp dead ahead — gun, crosshair and HUD framed like a screenshot. Kimi matches it: a monster down a textured hall, health, ammo, minimap. GLM ships a gorgeous 'HAZARD PROTOCOL' title screen with a working…
Voxel
Visual
GLM-5.2 9.0
·
Claude Opus 5 8.2
(+0.8)
· winner · flair
What I saw: GLM built the densest, most detailed city — windowed skyscrapers, a speed + coins HUD. Opus ran the furthest with the cleanest motion (Score 303). Kimi's runner plays fine but is unforgiving — it crashes within seconds.
Strengths & weaknesses I logged
Claude Opus 5
Strengths
- Frontier-class coding + agentic reasoning (Claude 5 family)
- 1M-token context — reads an entire codebase in one call
- Benched here with skill-infused game prompts the day of release
Trade-offs
- Premium pricing ($5/$25 per M) — route the everyday 90% to cheaper lanes
- Reasoning-by-default eats token budgets unless tuned per call
GLM-5.2
Strengths
- 1M-token context window — best-in-class long-document and large-codebase work
- Open weights — runs locally, no vendor lock-in, no token meter
- Top of the bench for cinematic visuals (neon city, synthwave, voxel runner)
Trade-offs
- Faceplanted on the Goldie Bench raycaster — the engine was great but it spawned the player inside a wall
- First-shot reliability lags Opus by a hair on consistency
Pricing & context — the spec sheet
| Spec | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| Vendor | Anthropic | Zhipu / Z.ai |
| Context window | 1,000,000 tokens | 1,000,000 tokens |
| Price | $5 / $25 per M | Open weights · free for individuals |
| Pricing detail | Anthropic's brand-new flagship — the first Opus of the Claude 5 family, with a 1M-token context window. Benched via API the day it dropped; game tasks use our skill-infused AAA build prompts. | Open-weights release: weights downloadable from Hugging Face for self-hosting, or runnable for free on z.ai for individuals (commercial use has separate licensing). |
| Release | 2026-07 | 2026-06-14 |
| Bench coverage | 50/50 scored · avg 8.27/10 | 47/47 scored · avg 7.77/10 |
The verdict — which should you pick?
Across 47 scored shared tasks, Claude Opus 5 averaged 8.29/10, beating GLM-5.2's 7.77/10 by 0.52 points. Pick Claude Opus 5 when the build has to ship on the first prompt and you can afford the trade-offs in the comparison below.
If you only run one of these inside your stack, the head-to-head average above is the call. If you can run both, my honest play is to wire Claude Opus 5 and GLM-5.2 both into the Agent Operating System and dispatch each from the kanban by task type — hardest agentic builds → Claude Opus 5, long-context agent loops — pasting a whole codebase into one prompt → GLM-5.2. That's the same setup I run for the 4,000+ founders inside the AI Profit Boardroom.
FAQ — Claude Opus 5 vs GLM-5.2
Which is better, Claude Opus 5 or GLM-5.2?
On Goldie Bench, Claude Opus 5 averages 8.29/10 across the shared tasks, with 13 gold, 7 silver, 5 bronze overall. GLM-5.2 averages 7.77/10, with 5 gold, 0 silver, 0 bronze. Claude Opus 5 wins the head-to-head 36–10.
How much does Claude Opus 5 cost vs GLM-5.2?
Claude Opus 5: Anthropic's brand-new flagship — the first Opus of the Claude 5 family, with a 1M-token context window. Benched via API the day it dropped; game tasks use our skill-infused AAA build prompts. GLM-5.2: Open-weights release: weights downloadable from Hugging Face for self-hosting, or runnable for free on z.ai for individuals (commercial use has separate licensing).
What's the context window for Claude Opus 5 vs GLM-5.2?
Claude Opus 5 has a 1,000,000 tokens context window. GLM-5.2 has a 1,000,000 tokens context window.
When should I pick Claude Opus 5 over GLM-5.2?
Pick Claude Opus 5 for: Hardest agentic builds; Whole-repo reasoning; Frontier one-shots. The trade-off is the weaknesses we logged on the bench: Premium pricing ($5/$25 per M) — route the everyday 90% to cheaper lanes; Reasoning-by-default eats token budgets unless tuned per call.
When should I pick GLM-5.2 over Claude Opus 5?
Pick GLM-5.2 for: Long-context agent loops — pasting a whole codebase into one prompt; Cinematic visual builds — landing pages, voxel scenes, synthwave runners; Anyone who needs to run a frontier coder locally for $0. The trade-off is the weaknesses we logged on the bench: Faceplanted on the {{SITE_NAME}} raycaster — the engine was great but it spawned the player inside a wall; First-shot reliability lags Opus by a hair on consistency.
How does Goldie Bench score Claude Opus 5 vs GLM-5.2?
Every demo on this page was built by Julian Goldie inside the Agent Operating System — same fixed prompt for both models, one shot, single HTML file out. Each result gets a 0–10 score on whether it ran, how close it hit the brief, and how good it looked. The highest score on each task gets gold; second gets silver; third gets bronze. See methodology for full provenance.
Related comparisons
Other head-to-heads using the same scoring system:
Claude Opus 5 vs Fusion GLM-5.2 vs Fusion Claude Opus 5 vs Hermes MoA GLM-5.2 vs Hermes MoA Claude Opus 5 vs GPT-5.6 Sol GLM-5.2 vs GPT-5.6 Sol Claude Opus 5 vs Claude Fable 5 GLM-5.2 vs Claude Fable 5Full model pages: Claude Opus 5 · GLM-5.2 · back to the leaderboard
The same stack Julian uses
Run this stack yourself.
Every demo on this bench was built inside the Agent Operating System — one prompt, one shot, single HTML file out. The Agent OS, the prompts, the templates, the weekly walkthroughs and 4,000+ founders shipping with it every day all live inside the AI Profit Boardroom.
4,000+founders
258documented wins
38countries
$59/momonthly