BuilderBenchby the LAB
26 September 2026 · v2 · pack 2.0.0
4-hour build window · Extra High (xhigh) · Standard
the reveal · personal assessment with accepted recording limitations
One owner-scored personal assessment of one GPT-6 Astra run with Codex on the v2 pack. This is not a universal model ranking or a statistical claim.
Codex harnessfinal score
53.75/100
API estimate in US dollars · see coverage below · 20 of 31 records priced
API equivalent · not a bill
personal assessment with accepted recording limitations. one run per setup. the limits are below.
quality is the weighted average of the task scores. autonomy is how little owner rescue the setup needed. proactivity is useful initiative inside the brief. each is scored out of 100.
the final score is quality × 70% + autonomy × 20% + proactivity × 10%. cost never buys points.
dated references
each bar is one owner review of one run. the packs, tasks and evidence limits differ, and each bar says which pack and which kind of review it was. scores only compare inside one pack.
task breakdown
each task is scored out of 100. its weight is its share of the quality score.
| task | GPT-6 Astrascore /100 |
|---|---|
| 01 · Writing15% of quality | 32 |
| 02 · Video editing15% of quality | 43.33 |
| 03 · Landing page15% of quality | 35 |
| 04 · Motion5% of quality | 20 |
| 05 · Browser + computer15% of quality | 90 |
| 06 · Research10% of quality | 50 |
| 07 · Automation10% of quality | 80 |
| 08 · Founder conversations5% of quality | 65 |
| 09 · Thumbnails10% of quality | 52 |
| task | GPT-6 AstraAPI-equivalent estimate |
|---|---|
| Writing | US$2.90recorded usage |
| Video editing | US$28.81partial usage |
| Landing page | US$15.25recorded usage |
| Motion | US$16.32recorded usage |
| Browser + computer | US$7.55recorded usage |
| Research | US$3.16recorded usage |
| Automation | US$2.47recorded usage |
| Founder conversations | US$0.89recorded usage |
| Thumbnails | US$10.15partial usage |
estimates, not bills. “unavailable” means no usage was recorded for that task. it does not mean the task was free.
API pricing
the rates are what the provider listed on the date shown. the estimate prices the usage this run recorded. the bill is what was actually charged. they are kept apart on purpose.
no rate card was frozen with this result.this result was published before rate cards were required. its recorded estimate keeps the price basis it had on the day. no newer price has been applied to it.
API-equivalent estimates price recorded execution usage at published API rates. They are not bills. Missing native usage, setup, calibration, coordination, grading, presentation, subscription fees and unreceipted services are excluded. Actual incremental spend is unknown.
review conditions
The owner rated the saved outputs of each of the nine v2 tasks, plus autonomy and proactivity, then locked the scores. An accepted evidence limitation covers the browser and computer task. Quality is the weighted task score. v1 to v3 use 70% quality, 20% autonomy and 10% proactivity.
frozen setup
exact model and harness version strings are not in this public export, so they stay unknown here.
limits
p.s.
live count, 24 oct 2026.