Lennox Saint
Open menu

BuilderBenchby the LAB

24 September 2026 · v1 pilot · pack 1.0.0
4-hour build window · Extra High (xhigh) · Standard

full result · one owner-scored pilot

Opus 5.5 scored 68.26. GPT-6 Sol scored 45.55.

One owner-scored pilot of each model, harness and tool setup. This is not a universal model ranking or a statistical claim.

words used on this page, in plain English
test (pack)
a fixed set of tasks. every AI in the same pack gets the same tasks, so scores only compare inside one pack.
score out of 100
70% how good the finished work was (quality), 20% how little help it needed (autonomy), 10% useful extras nobody asked for (proactivity).
harness
the app or tool that runs the AI, like Codex or Claude Code. the score belongs to the AI and its harness together.
cleared result
a result Lennox checked and approved for publishing.
owner review
Lennox reviewed and scored the work. some checks are automated, and on one result he asked Claude to rate some parts for him. each result page says which. each result is one review of one run per AI.
named or blind review
named: Lennox could see which AI made the work. blind: the names were hidden until the scores were saved.
partial review
some of the work could not be fully checked. the result page says which parts.
API estimate
what the run would have cost at the AI company's listed prices. it is an estimate, not what was paid.
reasoning (xhigh)
how hard the AI was set to think. xhigh means extra high, the highest setting used.

final scores out of 100

full result · BuilderBench test of 24 September 2026 · 2 AIs on this page

Claude harnesswon this test

Opus 5.5

68.26/100

quality 70% of the score
61.80
autonomy 20%
80
proactivity 10%
90
US$49.67

API estimate in US dollars · see coverage below
API equivalent · not a bill

full result · BuilderBench test of 24 September 2026 · 2 AIs on this page

Codex harnessfinal score

GPT-6 Sol

45.55/100

quality 70% of the score
46.50
autonomy 20%
40
proactivity 10%
50
US$12.75

API estimate in US dollars · see coverage below
API equivalent · not a bill

Opus 5.5 leads by 22.71 points.

one owner-scored pilot. one run per setup. winning one test does not make an AI the best overall. the limits are below.

how each score is built

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.

  • Opus 5.5: 43.26 + 16.00 + 9.00 = 68.26
  • GPT-6 Sol: 32.55 + 8.00 + 5.00 = 45.55

dated references

every cleared score, by date. not a leaderboard.

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.

  1. Opus 5.524 Sep 2026 · v1 · pack 1.0.0 · with Claude
    one owner-scored pilot · this result
    68.26
  2. GPT-6 Sol24 Sep 2026 · v1 · pack 1.0.0 · with Codex
    one owner-scored pilot · this result
    45.55
  3. Gemini 4 Neon19 Nov 2026 · v3 · pack 3.1.0 · with BuilderBench API harness
    owner-finalized partial personal review, with 3 of 28 ratings made by the bench's conductor (claude opus 5.5) and disclosed
    58.76
  4. Gemini 4 Hydrogen12 Nov 2026 · v3 · pack 3.1.0 · with BuilderBench API harness
    owner review, with 7 of 28 ratings made by claude and disclosed
    51.11
  5. Gemini 4 Carbon11 Nov 2026 · v3 · pack 3.1.0 · with BuilderBench API harness
    owner-finalized partial personal assessment
    49.04
  6. Opus 5.510 Nov 2026 · v3 · pack 3.1.0 · with Claude
    owner-finalized partial personal assessment
    74.25
  7. Fable 5.210 Nov 2026 · v3 · pack 3.1.0 · with Claude
    owner-finalized partial personal assessment
    69.43
  8. GPT-6.1 Terra27 Oct 2026 · v3 · pack 3.0.0 · with Codex
    owner-finalized partial personal review
    48.88
  9. Kimi K3.1 Pro27 Oct 2026 · v3 · pack 3.0.0 · with BuilderBench API harness (OpenRouter)
    owner-finalized partial personal review
    46.68
  10. Gemini 4 Argon14 Oct 2026 · v3 · pack 3.0.0 · with BuilderBench API harness
    owner-finalized partial personal review
    45.38
  11. Muse Wonder Full14 Oct 2026 · v3 · pack 3.0.0 · with Muse Code
    owner-finalized partial personal review
    39.91
  12. GPT-6 Luna8 Oct 2026 · v3 · pack 3.0.0 · with Codex
    owner-finalized partial personal assessment
    44.58
  13. Haiku 5.58 Oct 2026 · v3 · pack 3.0.0 · with Claude
    owner-finalized partial personal assessment
    39.18
  14. GPT-6.1 Sol30 Sep 2026 · v3 · pack 3.0.0 · with Codex
    owner-finalized partial personal assessment
    41.75
  15. Sonnet 5.529 Sep 2026 · v3 · pack 3.0.0 · with Claude
    owner-finalized partial personal assessment
    48.98
  16. GPT-6 Astra26 Sep 2026 · v2 · pack 2.0.0 · with Codex
    personal assessment with accepted recording limitations
    53.75

task breakdown

where each setup gained and lost ground

each task is scored out of 100. its weight is its share of the quality score.

swipe the table sideways to see every column.

Task scores out of 100 for each setup, with each task’s share of the quality score
taskOpus 5.5score /100GPT-6 Solscore /100
01 · Writing20% of quality4930
02 · Video editing15% of quality4010
03 · Frontend design15% of quality6040
04 · Motion graphics10% of quality6030
05 · Browser & computer15% of quality7090
06 · Research & decisions10% of quality7055
07 · Intake automation10% of quality9590
08 · Founder conversation5% of quality8040

API pricing

rates, the run estimate and the bill are three different things.

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.

Opus 5.5

official API rates

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.

recorded estimate for this run
US$49.67Provider-reported API estimates for the recorded task rows.
actual bill
unknown. billing for this run: existing subscriptions. no receipted charge exists for it, and unknown is not zero.

GPT-6 Sol

official API rates

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.

recorded estimate for this run
US$12.75Partial recorded usage only; unknown attempts and shared overhead are excluded.
actual bill
unknown. billing for this run: existing subscriptions. no receipted charge exists for it, and unknown is not zero.

API-equivalent estimates are not bills. Actual incremental spend is unknown. The methods and coverage differ, so these figures do not support an exact cost ratio.

correction

the correction is a receipt

2026-09-24 15:13 UTC · fixed before this result went public

what happened: after the scores were locked, live-chat deliverables were still flagged as missing.

why: stale missing-file flags from the offline window. the frozen files and their hashes were checked again, and they were there.

what did not change: my ratings of the work did not change. only the missing-file flags did.

Scores before and after the correction
setupscorebeforeafter
Opus 5.5 · Claudequality57.8061.80
Opus 5.5 · Claudeoverall65.4668.26
GPT-6 Sol · Codexquality44.5046.50
GPT-6 Sol · Codexoverall44.1545.55

review conditions

how this run was reviewed

The owner scored quality, autonomy and proactivity after a blind review. Quality is the weighted score across eight task families. The score lock happened before identities were revealed.

scores locked
yes
names revealed
yes, after lock

frozen setup

what was tested

pack version
1.0.0
build window
4 hours
reasoning
Extra High (xhigh)
speed
Standard
billing
Existing subscriptions
scored runs
1 per setup

exact model and harness version strings are not in this public export, so they stay unknown here.

limits

what this result does not prove

  • This was one pilot scored by one owner, not a universal ranking.
  • The score belongs to each full setup: model, harness, tools and settings.
  • The two cost estimates use different methods and coverage.
  • Actual incremental spend remains unknown because the run used existing subscriptions.
  • Complete normal-speed watched-and-listened review of both full video outputs was not established.
BuilderBench pilot result: Opus 5.5 scored 68.26 and GPT-6 Sol scored 45.55, with API-equivalent estimate caveats.
the result graphic. one owner-scored pilot. costs are estimates, not bills. every number in it is also on this page as text.

p.s.

get the next BuilderBench result first.

  • the exact prompt i typed
  • what came back
  • where it fell over

free. unsubscribe any time, one click. privacy.

  • 34,473 on Threads
  • 3,612 readers
  • 1.91K on YouTube

live count, 22 nov 2026.