xAI

Grok

Snappy + real-time — the X-native model.

Context256,000 tokens

PricingSubscription via X Premium

Tasks tested47

Avg score8.09/10 average

Medals🥇5 🥈1 🥉1

Release2026-04

Official sitegrok.com ↗

Official vendor source

Grok is built by xAI — see the vendor's own product page, pricing, and docs at grok.com.

Visit grok.com →

What is Grok?

Grok is the xAI frontier model with a 256,000 tokens context window, released 2026-04. Tagline: Snappy + real-time — the X-native model.. Official source: grok.com.

Pricing detail. Bundled with X (Twitter) Premium subscription — no per-token bill for end users, no individual API pricing for the chat product.

How I use it inside the Agent OS. Used for real-time content workflows where the model needs current X timeline context. Standalone bench scoring pending.

What I built with Grok

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 Grok shipped on the bench: 47 one-shot demos across 256,000 tokens of context. Of those, 43 are scored against the field with my honest 0–10 from the source guides at agentos.guide.

Strengths

Real-time access to X timeline data — unique signal no other model has
Snappy latency on shorter prompts
256K context window keeps pace with the open-weights field

Trade-offs

13 demos on the bench but zero have curated 0–10 verdicts yet — currently unranked
API access is gated behind X Premium, awkward for backend agent loops

Best for

Workflows that need live X / Twitter context
Snappy prompts where latency matters
Researchers comparing X-native models against the rest of the field

Every benchmark — Grok's full scorecard

All 43 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 Grok deep dive →.

Every demo by Grok

47 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.

Grok one-shot build of Crypt — GoldieBench AI benchmark screenshot

Grok one-shot build of Dragonrealm — GoldieBench AI benchmark screenshot

Doom-style FPS with sprite enemies, gun + muzzle flash + ammo/health HUD, textures, pointer-lock mouse-look. 22KB.

Grok one-shot build of Racing — GoldieBench AI benchmark screenshot

3D arcade racer, third-person, banking turns, drift mechanic, obstacles, lap timer. 29KB on second retry.

Grok one-shot build of Skyrim — GoldieBench AI benchmark screenshot

Grok one-shot build of Twilightvale — GoldieBench AI benchmark screenshot

Twilightvale 🥇

Twilight Vale — 3D open-world RPG with hand-crafted village, NPCs, combat, day/night, weather, inventory. 38KB — densest build of the bench, edges out Fusion's 32KB.

Grok one-shot build of Voxelcraft — GoldieBench AI benchmark screenshot

Minecraft-style voxel sandbox, fly + break + place, day/night cycle, hotbar. 22KB.

Grok one-shot build of Gtafoot — GoldieBench AI benchmark screenshot

10KB · plays clean · three, webgl, input (re-rolled)

Grok one-shot build of Gtadrive — GoldieBench AI benchmark screenshot

11KB · plays clean · three, webgl, input

Grok one-shot build of Aipbpromo — GoldieBench AI benchmark screenshot

11KB · plays clean · plain

Grok one-shot build of Parachute — GoldieBench AI benchmark screenshot

9KB · plays clean · three, webgl, input, rAF

Grok one-shot build of Flightsim — GoldieBench AI benchmark screenshot

9KB · plays clean · three, webgl, input, rAF

Grok one-shot build of Arcade — GoldieBench AI benchmark screenshot

A crisp neon Asteroids — glowing ship, lives, score, drifting rocks, real sound. Shippable straight out of one prompt.

Grok one-shot build of Dogfight — GoldieBench AI benchmark screenshot

3D dogfight game with fighter jet, enemies, missiles + guns, sky + ground, crosshair + speed + altitude HUD. 30KB.

Grok one-shot build of Dragonflight — GoldieBench AI benchmark screenshot

Dragonflight 🥉

Fly a dragon through neon rings with score + fire-breath + fury HUD. 28KB.

Grok one-shot build of Game — GoldieBench AI benchmark screenshot

Open-ended 'make a game' — Grok shipped a juicy 28KB build with score HUD, lives, sound, polish.

Grok one-shot build of Neonblaster — GoldieBench AI benchmark screenshot

Juicy 2D arcade space shooter — waves, bosses, power-ups, screen-shake, synth music via Web Audio, neon particles. 24KB.

Grok one-shot build of Neoncity — GoldieBench AI benchmark screenshot

Cyberpunk neon city flythrough with light trails, holograms, fog. Drag to look around. 19KB.

Grok one-shot build of Neonracer — GoldieBench AI benchmark screenshot

Top-down neon racer with vapor trails, synthwave aesthetic, drift, lap timer, procedural track. 22KB.

Grok one-shot build of Outrun — GoldieBench AI benchmark screenshot

Pseudo-3D OutRun racer — curving road, synthwave sunset, parallax scenery. Arrow-key steering. 24KB.

Grok one-shot build of Pool — GoldieBench AI benchmark screenshot

Canvas2D billiards with 16 balls, pockets, click-drag aim + power. 16KB.

Grok one-shot build of Raycaster — GoldieBench AI benchmark screenshot

Canvas 2D raycaster maze with WASD + mouse-look, floor/ceiling, distance fog, weapon bob. 20KB.

Grok one-shot build of Rpg — GoldieBench AI benchmark screenshot

35KB top-down RPG with tilemap, walkable terrain, NPCs, combat, HP/MP UI, inventory. Beats Fusion's lighter 26KB attempt on density.

Grok one-shot build of Nordiccrypt — GoldieBench AI benchmark screenshot

Grok one-shot build of Landing — GoldieBench AI benchmark screenshot

A genuinely premium keynote page: clean nav, a gradient headline, dual buttons, tasteful type. From one sentence. Grok Build's best work of the lot.

Grok one-shot build of Webos — GoldieBench AI benchmark screenshot

Web-OS desktop with wallpaper, dock, draggable resizable windows for Notes/Paint/Terminal/Calculator. 33KB — beats Fusion's 24KB attempt on density.

Grok one-shot build of Blackhole — GoldieBench AI benchmark screenshot

A striking black event horizon ringed by a white-hot accretion disk over a real starfield. The sharper sentence asked for a brighter ring and a defensive render — and got both.

Grok one-shot build of Boids — GoldieBench AI benchmark screenshot

Hundreds of glowing fish with flowing trails and the lone red predator they flee. The prettiest of the pure-visual set.

Grok one-shot build of Cloth — GoldieBench AI benchmark screenshot

Verlet-integrated cloth simulation, pin corners, drag to deform. 14KB.

Grok one-shot build of Fluid — GoldieBench AI benchmark screenshot

A round bowl of ~500 distinct glowing droplets that fall, pool at the bottom and slosh when you tilt with the mouse, with colour themes. The first version melted into one white blob; this one keeps every droplet separate, like real liquid.

Grok one-shot build of Fractal — GoldieBench AI benchmark screenshot

A clean, vivid Mandelbrot — the classic shape in smooth rainbow colour, with an auto-zoom and scroll-to-zoom. Renders reliably now.

Grok one-shot build of Galaxy — GoldieBench AI benchmark screenshot

A bright spiral with a hot-white core and clear arms fading from gold to cyan. Far brighter and more obviously a galaxy than the dim first attempt.

Grok one-shot build of Orbit — GoldieBench AI benchmark screenshot

A proper inner solar system now — a glowing Sun, four planets riding clean elliptical rings, a starfield, a data HUD and play/speed controls. The vague first sentence drew blurry circles; this one reads instantly.

Grok one-shot build of Particleforge — GoldieBench AI benchmark screenshot

Sculpt swirling particle systems with mouse gravity, multiple presets, FPS counter. 18KB.

Grok one-shot build of Pathtracer — GoldieBench AI benchmark screenshot

WebGL fragment-shader path tracer with Cornell-box scene, soft shadows, sample accumulation. 18KB.

Grok one-shot build of Reactiondiff — GoldieBench AI benchmark screenshot

Gray-Scott reaction-diffusion on WebGL with click-to-seed + f/k sliders. 19KB.

Grok one-shot build of Solar — GoldieBench AI benchmark screenshot

Solar system on canvas with orbits + Saturn rings + moons. 12KB — compact build, drag-to-orbit + scroll-to-zoom. Slightly less polish than Fusion's glass UI version.

Grok one-shot build of Wormhole — GoldieBench AI benchmark screenshot

Three.js wormhole tunnel with distorted starfield, hold-space-to-accelerate. 18KB.

Grok one-shot build of Aurora — GoldieBench AI benchmark screenshot

Aurora ribbons over mountain ridge with stars. Simpler build (9KB) — lighter on detail than Fusion.

Grok one-shot build of Fireworks — GoldieBench AI benchmark screenshot

Click-to-launch fireworks with particle trails, sparkle physics. 14KB. No audio (vs Fusion's whoosh+boom).

Grok one-shot build of Lavalamp — GoldieBench AI benchmark screenshot

The capsule's full of bright gooey blobs now — cyan, pink and white metaballs rising and merging. The raymarched-shader version came up empty twice, so I had it rebuilt in plain Canvas 2D, and now it renders every time.

Grok one-shot build of Matrix — GoldieBench AI benchmark screenshot

Classic matrix rain. Smaller build (6KB) — straightforward implementation, on brief.

Grok one-shot build of Plasma — GoldieBench AI benchmark screenshot

Full-screen plasma with palette switcher + click-to-ripple. Lighter build (9KB) than Fusion's.

Grok one-shot build of Synthwave — GoldieBench AI benchmark screenshot

A scanline sun behind a bold silhouette, a neon grid running to the horizon. Proper retro atmosphere, zero clicks.

Grok one-shot build of Terrain — GoldieBench AI benchmark screenshot

Tron procedural terrain flythrough with glowing grid mountains. Drag to steer. 18KB.

Grok one-shot build of Voxel — GoldieBench AI benchmark screenshot

A colourful 3D voxel city with a score and coins HUD and a polished game-over card. Like every runner it ends fast — but the build is excellent.

Grok one-shot build of Waves — GoldieBench AI benchmark screenshot

Gerstner ocean waves with sun reflection + foam crests. Drag to look around. 12KB.

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

Compare Grok against every other model

Every head-to-head featuring Grok. Verdicts shown for scored pairs.

Fusion leads 31–2

Grok vs Claude Opus 5

Claude Opus 5 leads 28–15

Grok vs Hermes MoA

Hermes MoA leads 27–14

Grok vs GPT-5.6 Sol

GPT-5.6 Sol leads 28–14

Grok vs Claude Fable 5

Claude Fable 5 leads 20–17

Grok vs Qwen 3.8

Qwen 3.8 leads 28–10

Grok vs MiniMax M3

Grok leads 19–7

Grok vs Fugu Ultra

Fugu Ultra leads 16–14

Grok vs Kimi K3

Kimi K3 leads 25–18

Grok vs GLM-5.2

Grok leads 24–9

Grok vs Fugu Mini

Grok leads 14–10

Grok vs Opus 4.8

Grok leads 22–8

Grok vs Kimi K2.7

Grok leads 13–2

Grok vs Qwable 5 27B Coder

Grok leads 25–10

Grok vs Gemini 3.6 Flash

Grok leads 27–16

Grok vs Claude Sonnet 5

Grok leads 26–16

Grok vs Qwen 3.7

Grok leads 34–1

Grok vs Fugu Ultra 1.1

Grok leads 14–4

Grok vs Inkling

Grok leads 39–4

Grok vs Agents-A1

Grok leads 38–0

Grok vs Gemma 4 12B · MLX

Grok leads 38–0

Grok vs Laguna XS 2.1

Grok leads 38–0

Grok vs Qwythos 9B

Grok leads 38–0

Grok vs LongCat-2.0

Grok leads 1–0

Grok leads 5–1

Grok vs Gemma-4 12B Coder

Grok leads 6–0

Grok vs DeepSeek V4 Flash

47 shared tasks · unscored

Grok vs DeepSeek V4 Pro

47 shared tasks · unscored

Grok vs Kimi K2.7 · Fast

47 shared tasks · unscored

Grok vs Kimi K2.7 · No-Think

47 shared tasks · unscored

Grok vs Kimi K2.7 · Quality

47 shared tasks · unscored

Grok vs Ornith 1.0

42 shared tasks · unscored

Grok vs Claude Mythos 5

Grok vs Kilo Code

See all 66 comparisons across every model →

Quick pill index

Direct comparisons against every other scored model on the bench:

Grok vs Fusion Grok vs Claude Opus 5 Grok vs Hermes MoA Grok vs GPT-5.6 Sol Grok vs Claude Fable 5 Grok vs Qwen 3.8 Grok vs MiniMax M3 Grok vs Fugu Ultra Grok vs Kimi K3 Grok vs GLM-5.2 Grok vs Fugu Mini Grok vs Opus 4.8 Grok vs Kimi K2.7 Grok vs Qwable 5 27B Coder Grok vs Gemini 3.6 Flash Grok vs Claude Sonnet 5 Grok vs Qwen 3.7 Grok vs Fugu Ultra 1.1 Grok vs Inkling Grok vs Agents-A1 Grok vs Gemma 4 12B · MLX Grok vs Laguna XS 2.1 Grok vs Qwythos 9B Grok vs LongCat-2.0 Grok vs Hy3 Grok vs Gemma-4 12B Coder

Read more on agentos.guide: /grok-build

Grok — frequently asked

What is Grok?

Grok is xAI's AI model — Snappy + real-time — the X-native model. It has a 256K tokens context window and was released 2026-04.

How good is Grok at coding and one-shot builds?

On the GoldieBench one-shot build benchmark it averages 8.09/10 across 43 scored tasks, with 5 gold, 1 silver and 1 bronze medals.

How much does Grok cost?

Subscription via X Premium. Bundled with X (Twitter) Premium subscription — no per-token bill for end users, no individual API pricing for the chat product.

Where can I see Grok 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 →