Get the Agent OS + join 4,000+ founders inside the AI Profit Boardroom → Join AIPB ($59/mo)

GoldieBench deep dive

Grok's full benchmark breakdown.

Snappy + real-time — the X-native model. Every number below is either one of my own judged one-shot builds - playable on this site - or an outside result with its source linked. Nothing display-only, nothing vibes.

Data refreshed
2026-07-17
Scored tasks
43
Reading time
9 min
External sources
4

01 · The headline numbers

GoldieBench average
8.09/10
43 scored one-shot tasks
Board rank
#6
of 17 ranked frontier models
Task medals
5🥇 1🥈 5🥉
outright wins on shared briefs
Context
256,000 tokens
Subscription via X Premium

02 · Every benchmark, bar by bar

All 43 scored tasks, best first. Hover for the judge's comment; click through for every model's build on that task. Or overlay other models in the interactive graphs.

03 · Where it wins - with the judge's own words

“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.”

- the judge's verdict, unedited

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

- the judge's verdict, unedited

“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.”

- the judge's verdict, unedited

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

- the judge's verdict, unedited

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

- the judge's verdict, unedited

04 · Where it struggles - quoted, not hidden

Boards that hide the weak rows are brochures. These are Grok's lowest scored builds, verdicts unedited:

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

- the judge's verdict, unedited

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

- the judge's verdict, unedited

“11KB · plays clean · plain”

- the judge's verdict, unedited

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

- the judge's verdict, unedited

05 · Category breakdown vs the whole field

Games (19 tasks)

Grok8.3
field avg7.1

Others (0 tasks)

field avg7.4

Pages (3 tasks)

Grok8.3
field avg7.5

Sims (12 tasks)

Grok8.0
field avg6.7

Visuals (9 tasks)

Grok7.7
field avg6.7

06 · Outside signals - every row sourced

My bench measures one thing: judged one-shot builds. These outside rows measure other things - kept separate, never blended into the GoldieBench average, every value linked to where it comes from.

Vendor-published numbers

BenchmarkResultSource
SWE-bench Pro64.7%x.ai
Terminal-Bench 2.183.3%x.ai
DeepSWE 1.062.0%x.ai
SWE Marathon (pass@1)29.0% - ahead of Opus 4.8 (26.0%) and Fable max (24.0%)x.ai
Token efficiency15,954 avg output tokens per SWE-bench Pro task vs 67,020 for Opus 4.8 max - a 4.2x gapx.ai

Independent evaluations

BenchmarkResultSource
BenchLM weighted board6th overall (76.97) - 6th coding (71.25) - 12th agentic (62.09), Supported evidencebenchlm.ai

07 · Head-to-head records

08 · Methodology + honest limits

Every GoldieBench score is a one-shot, single-file build from an identical prompt - no retries, no hand-fixing - rendered for real, screenshotted, and scored 0-10 by one judge model on one rubric across the whole field. Failures score as failures. What this bench does NOT measure: multi-turn agent work, long-context recall, or API latency - that is what the sourced outside rows are for. Full method: /methodology.

Frequently asked questions

01How does Grok perform on GoldieBench?

Grok averages 8.09/10 across 43 scored one-shot build tasks, ranking #6 of 17 ranked frontier models, with 5 task golds, 1 silvers and 5 bronzes. Every score is a real judged build you can open and play on this site.

02What is Grok best at?

Its strongest scored build is Twilightvale at 9.5/10. By category it averages Game 8.3, Page 8.3, Sim 8.0, Visual 7.7 on the bench.

03How much does Grok cost?

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

Source ledger

  1. 01Official vendor site: grok.comgrok.com
  2. 02SWE-bench Prox.ai
  3. 03BenchLM weighted boardbenchlm.ai
  4. 04Julian's guide: /grok-buildagentos.guide

Every model's breakdown

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