Anthropic

Claude Opus 5

The new Anthropic flagship — benched on all 45 one-shot builds the day it landed.

Context1,000,000 tokens

Pricing$5 / $25 per M

Tasks tested22

Avg score5.60/10 average

Medals🥇0 🥈0 🥉1

Release2026-07

Official siteanthropic.com/claude ↗

Official vendor source

Claude Opus 5 is built by Anthropic — see the vendor's own product page, pricing, and docs at anthropic.com/claude.

Visit anthropic.com/claude →

What is Claude Opus 5?

Claude Opus 5 is the Anthropic frontier model with a 1,000,000 tokens context window, released 2026-07. Tagline: The new Anthropic flagship — benched on all 45 one-shot builds the day it landed.. Official source: anthropic.com/claude.

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.

How I use it inside the Agent OS. Benched on all 45 GoldieBench tasks via API on release day, incremental deploys as scores landed.

What I built with Claude Opus 5

Every model on Goldie Bench gets the same fixed prompt set — one shot, single HTML file out — and I score the result 0–10 inside the Agent Operating System. Here's what Claude Opus 5 shipped on the bench: 22 one-shot demos across 1,000,000 tokens of context. Of those, 13 are scored against the field with my honest 0–10 from the source guides at agentos.guide.

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

Best for

Hardest agentic builds
Whole-repo reasoning
Frontier one-shots

Every benchmark — Claude Opus 5's full scorecard

All 13 scored tasks, best first — the judge's 0–10 on the same rubric as the whole field. Click any bar for that task's cross-model page, or open this scorecard in the interactive graphs. Full editorial breakdown with judge quotes and sourced outside research: the Claude Opus 5 deep dive →.

GtadriveGame8.4 🥉DogfightGame8.4 DragonrealmGame8.3 DragonflightGame8.2 NeoncityGame7.4 DoomGame6.8 NeonblasterGame6.8 ArcadeGame6.5 FlightsimGame3.0 GtafootGame2.5 GameGame2.5 CryptGame2.0 NeonracerGame2.0

Every demo by Claude Opus 5

22 live demos, sorted by category. Click any tile to play the actual one-shot result. Verdicts and 0–10 scores are pulled from the source guides where I posted them publicly.

Claude Opus 5 one-shot build of Crypt — GoldieBench AI benchmark screenshot

The 3D scene fails to render — screen is blank except HUD, still showing 'KINDLING THE TORCH…' boot text. The source contains syntax errors (e.g. `0xffa martial=0`, `0x built=0`) that break the script, so no dungeon, torches, or enemies ever appear.

Claude Opus 5 one-shot build of Dragonrealm — GoldieBench AI benchmark screenshot

Gorgeous atmospheric frozen world — aurora, moon glow, snowy terrain, low-poly pines, ruins, and a well-modeled armored third-person character with sword DRAWN and full HUD (compass, health/stam, kill counter). Strong and clearly shippable, but no visible enemies/dragon or combat action on screen keeps it just below the field's best; feels like a polished walking sim awaiting foes.

Claude Opus 5 one-shot build of Doom — GoldieBench AI benchmark screenshot

Strong polished HUD, textured raycaster walls, weapon viewmodel, and a detailed minimap show real engine work, plus counters imply combat (01/10 kills, 09 demons left). But the screenshot shows NO visible enemies on screen — combat can't be confirmed in the frame, so it reads closer to a walking sim than the enemy-chase brief demands.

Claude Opus 5 one-shot build of Gtafoot — GoldieBench AI benchmark screenshot

The HUD (health/stamina bars, ammo, wanted stars, minimap, controls) renders but the entire 3D scene is a blank flat purple void — no city, no hero, no pedestrians, no enemies visible, indicating the WebGL render failed or nothing is drawn. Despite ambitious source code, the actual screenshot shows a non-rendering game world.

Claude Opus 5 one-shot build of Gtadrive — GoldieBench AI benchmark screenshot

Gorgeous cohesive art direction — detailed low-poly car with wheels/shadow, layered cityscape, refined HUD with integrity/velocity/wanted stars and a live minimap showing cop (red) and player markers plus traffic (purple car visible). Pursuit=2 and full control set (fire/steal/boost) signal working combat, but no cops or firing visible in the frame keeps it just short of the very top.

Claude Opus 5 one-shot build of Flightsim — GoldieBench AI benchmark screenshot

The HUD is elaborate and polished (airspeed, altitude, ADI, heading tape, throttle/kills panels), but the 3D scene renders as a blank tan void — no terrain, runway, aircraft, or enemies are visible, so the actual flight sim is effectively non-rendering.

Claude Opus 5 one-shot build of Arcade — GoldieBench AI benchmark screenshot

Visually impressive 3D neon arena with polished HUD, minimap, health/boost meters, combat feedback ('HULL HIT -9'), and hostiles on radar — but the brief explicitly asked for a CLASSIC arcade game (tetris/breakout/snake), and this is a 3D twin-stick shooter that ignores the pick entirely. Strong execution but off-brief.

Claude Opus 5 one-shot build of Dogfight — GoldieBench AI benchmark screenshot

Gorgeous detailed player jet, full flight HUD, functional radar showing 4 bogeys, terrain warning, and afterburner/boost systems all read as a shippable 3D dogfighter. Slightly held back because the screenshot shows no visible enemy aircraft in view or active combat/tracers, leaving the on-screen action a touch quiet versus the field's best.

Claude Opus 5 one-shot build of Dragonflight — GoldieBench AI benchmark screenshot

Gorgeous stylized dragon with articulated wings, spine and belly plates plus a polished full HUD (hull/fury/alt/spd/combo, rings, kills, hostiles, distance marker) — clearly shippable and combat-ready with 5 hostiles tracked. Held back from the top by no visible neon rings or enemies/fire-breath actually in frame, so the core ring-flight loop isn't proven on screen.

Claude Opus 5 one-shot build of Game — GoldieBench AI benchmark screenshot

The HUD renders beautifully (integrity/thrust/shield meters, score panel, crosshair, control hints), but the 3D scene is completely blank — no ground, mech, enemies, or landmarks visible, indicating the WebGL/three.js world failed to render. A source-code error (invalid `emissive:0x insertion` hex value in the material kit) likely broke script execution, leaving only the CSS HUD.

Claude Opus 5 one-shot build of Neonblaster — GoldieBench AI benchmark screenshot

Strong 3D presentation with polished HUD, minimap, ship model and space environment, but the brief asked for a 2D-style JUICY ARCADE space shooter with waves/bosses/power-ups — the screenshot shows a slow 3D cockpit-ish flight sim with faint asteroid-like triangles and no visible combat, projectiles, or enemy engagement (velocity 7 u/s, 0 kills). Reads more as an atmospheric flight demo than a punchy arcade blaster.

Claude Opus 5 one-shot build of Neoncity — GoldieBench AI benchmark screenshot

Strong polished HUD (hull/boost/speed/kills/hostile), a nicely modeled neon vehicle with soft shadow, and a coherent purple cyberpunk street with lane dashes and neon strips — but the scene reads sparse and dim, the neon-lit density is modest, and no enemies/combat are visible in-frame despite the combat-focused UI. Shippable and cohesive, but doesn't quite hit the vivid, dense neon-city bar to top the field.

Claude Opus 5 one-shot build of Neonracer — GoldieBench AI benchmark screenshot

The screenshot shows only raw JavaScript source text on a white page — nothing rendered, no canvas, no neon scene. Despite ambitious combat/vapor code in the source, the actual output is non-rendering.

every demo, in a grid · click any one to play

Compare Claude Opus 5 against every other model

Every head-to-head featuring Claude Opus 5. Verdicts shown for scored pairs.

Claude Opus 5 vs Fusion

Fusion leads 13–0

Claude Opus 5 vs Hermes MoA

Hermes MoA leads 11–2

Claude Opus 5 vs GPT-5.6 Sol

GPT-5.6 Sol leads 12–0

Claude Opus 5 vs Claude Fable 5

Claude Fable 5 leads 9–4

Claude Opus 5 vs Qwen 3.8

Qwen 3.8 leads 9–3

Claude Opus 5 vs Grok

Grok leads 10–1

Claude Opus 5 vs MiniMax M3

MiniMax M3 leads 12–1

Claude Opus 5 vs Fugu Ultra

Fugu Ultra leads 7–3

Claude Opus 5 vs Kimi K3

Kimi K3 leads 11–1

Claude Opus 5 vs GLM-5.2

GLM-5.2 leads 9–4

Claude Opus 5 vs Fugu Mini

Fugu Mini leads 5–3

Claude Opus 5 vs Opus 4.8

Opus 4.8 leads 9–4

Claude Opus 5 vs Kimi K2.7

Kimi K2.7 leads 7–1

Claude Opus 5 vs Qwable 5 27B Coder

Qwable 5 27B Coder leads 6–4

Claude Opus 5 vs Gemini 3.6 Flash

Gemini 3.6 Flash leads 5–4

Claude Opus 5 vs Claude Sonnet 5

Claude Sonnet 5 leads 9–4

Claude Opus 5 vs Qwen 3.7

Qwen 3.7 leads 9–4

Claude Opus 5 vs Inkling

Inkling leads 7–6

Claude Opus 5 vs Agents-A1

Claude Opus 5 leads 6–4

Claude Opus 5 vs Gemma 4 12B · MLX

Claude Opus 5 leads 7–3

Claude Opus 5 vs Laguna XS 2.1

Claude Opus 5 leads 7–3

Claude Opus 5 vs Qwythos 9B

Claude Opus 5 leads 7–2

Claude Opus 5 vs LongCat-2.0

LongCat-2.0 leads 2–0

Claude Opus 5 vs Hy3

Claude Opus 5 leads 3–2

Claude Opus 5 vs Gemma-4 12B Coder

Claude Opus 5 leads 1–0

Claude Opus 5 vs Kimi K2.7 · Fast

22 shared tasks · unscored

Claude Opus 5 vs Kimi K2.7 · No-Think

22 shared tasks · unscored

Claude Opus 5 vs Kimi K2.7 · Quality

22 shared tasks · unscored

Claude Opus 5 vs Ornith 1.0

18 shared tasks · unscored

Claude Opus 5 vs Fugu Ultra 1.1

Claude Opus 5 vs Claude Mythos 5

Claude Opus 5 vs Kilo Code

See all 66 comparisons across every model →

Quick pill index

Direct comparisons against every other scored model on the bench:

Claude Opus 5 vs Fusion Claude Opus 5 vs Hermes MoA Claude Opus 5 vs GPT-5.6 Sol Claude Opus 5 vs Claude Fable 5 Claude Opus 5 vs Qwen 3.8 Claude Opus 5 vs Grok Claude Opus 5 vs MiniMax M3 Claude Opus 5 vs Fugu Ultra Claude Opus 5 vs Kimi K3 Claude Opus 5 vs GLM-5.2 Claude Opus 5 vs Fugu Mini Claude Opus 5 vs Opus 4.8 Claude Opus 5 vs Kimi K2.7 Claude Opus 5 vs Qwable 5 27B Coder Claude Opus 5 vs Gemini 3.6 Flash Claude Opus 5 vs Claude Sonnet 5 Claude Opus 5 vs Qwen 3.7 Claude Opus 5 vs Inkling Claude Opus 5 vs Agents-A1 Claude Opus 5 vs Gemma 4 12B · MLX Claude Opus 5 vs Laguna XS 2.1 Claude Opus 5 vs Qwythos 9B Claude Opus 5 vs LongCat-2.0 Claude Opus 5 vs Hy3 Claude Opus 5 vs Gemma-4 12B Coder

Read more on agentos.guide:

Claude Opus 5 — frequently asked

What is Claude Opus 5?

Claude Opus 5 is Anthropic's AI model — The new Anthropic flagship — benched on all 45 one-shot builds the day it landed. It has a 1M tokens context window and was released 2026-07.

How good is Claude Opus 5 at coding and one-shot builds?

On the GoldieBench one-shot build benchmark it averages 5.60/10 across 13 scored tasks, with 0 gold, 0 silver and 1 bronze medals.

How much does Claude Opus 5 cost?

$5 / $25 per M. 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.

Where can I see Claude Opus 5 demos?

Every one-shot build is live and playable on this page and on the GoldieBench compare matrix — same prompt as every other model, no retries.

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

Join AIPB · $59/mo → Read the Agent OS guides →