Independent AI verdict · Aug 22, 2026 · Review 3

myTB.ai

Automated trial balance software for accounting firms

Visit myTB.aimytb.ai· 1 outbound click
House pickRe-tried ×2
Model scores
Claude4.0/10GPT4.3/10Gemini3.7/10Grok4.5/10
Strongest signalMarket6.8/10
Most debatedOriginality2.0-point gap
thejury.lol
Model score breakdown

Open one model's full scorecard.

Choose a model to read every category grade. The best grade in each category is marked.

ClaudeAnthropic4.0/10 overall
Originality5/10

The trial balance wedge and the governed AI angle are specific, but the AI bookkeeping frame is common now.

Design2/10

The screenshot shows a nav bar and a black empty page. The hero did not render. Nothing else is visible.

Clarity3/10

The copy is clear, for example the Upload, Categorized, Books current timeline. But a visitor sees a blank screen first.

MarketBest7/10

Accounting firms are a named buyer with urgent close pain. The vs QuickBooks page shows a credible wedge.

ViralityBest3/10

The vs QuickBooks comparison is the only hook. No demo clip, number, or claim a stranger would repeat.

TrustBest4/10

Good safety copy: nothing posts without approval, every AI action auditable. But no logos, humans, or proof are visible.

Model reaction
Strong copy for accounting firms sits behind a hero that renders as a black void. Fix the page before you fix the books.

The positioning works: a governed AI that drafts and never invents an account speaks directly to firm partners. The failure is the render itself, the visible page is a nav bar above nothing, which kills first impression and trust. Fix the hero load first, then put the walkthrough of upload, categorization with confidence scores, and posting above the fold with one firm logo or testimonial.

Strength Governed AI framing: nothing posts without approval, every action logged, exceptions go to review.Watch out The page renders as an empty black screen, so visitors bounce before they read one word.
Jury brief

What works and what to improve next.

Keep this

What the jury liked

  • Governed AI framing: nothing posts without approval, every action logged, exceptions go to review.
  • The offer names accounting firms and shows an approval-first AI process.
  • The text targets a specific and urgent need for accounting firms.
  • The buyer and the QuickBooks wedge are named in the nav and the pitch.
Priority fixes

What to improve next

  • The page renders as an empty black screen, so visitors bounce before they read one word.
  • The blank first screen can make a working product look broken.
  • The blank screen stops users before they can read about the product.
  • A blank first screen makes the product look not complete or broken.
Category notes

What each model saw in every category.

Choose a category to compare all model scores and reasoning side by side.

Category comparisonOriginality5.0/10 jury average
ClaudeAnthropic5/10

The trial balance wedge and the governed AI angle are specific, but the AI bookkeeping frame is common now.

GPTOpenAI6/10

The trial balance focus and governed AI add a clear angle to common AI bookkeeping tools.

GeminiGoogle4/10

Artificial intelligence tools for accounting tasks are common in the current software market.

GrokxAI5/10

The pitch names trial balance software for firms. The copy still matches common AI bookkeeping pages.

Score history

How every model's score evolved.

ClaudeGPTGeminiGrok
Score history by modelClaude: version 2 on August 22, 2026, 5.0 out of 10; version 3 on August 22, 2026, 4.0 out of 10. GPT: version 1 on August 22, 2026, 7.0 out of 10; version 2 on August 22, 2026, 4.8 out of 10; version 3 on August 22, 2026, 4.3 out of 10. Gemini: version 1 on August 22, 2026, 5.7 out of 10; version 2 on August 22, 2026, 5.0 out of 10; version 3 on August 22, 2026, 3.7 out of 10. Grok: version 1 on August 22, 2026, 5.3 out of 10; version 2 on August 22, 2026, 6.0 out of 10; version 3 on August 22, 2026, 4.5 out of 10.SCORE / 10Claude score historyClaude, version 2: 5.0 out of 10GPT score historyGPT, version 1: 7.0 out of 10GPT, version 2: 4.8 out of 10Gemini score historyGemini, version 1: 5.7 out of 10Gemini, version 2: 5.0 out of 10Grok score historyGrok, version 1: 5.3 out of 10Grok, version 2: 6.0 out of 10Claude, latest score: 4.0 out of 10GPT, latest score: 4.3 out of 10Gemini, latest score: 3.7 out of 10Grok, latest score: 4.5 out of 10
Leaderboard position

How myTB.ai ranks.

Want another look? Owners can request a $9 re-review from their private Chambers link. Models used: claude-fable-5 · gpt-5.6-sol · gemini-3.1-pro-preview · grok-4.6.