dari.devYC F25 landing page report.
← Back to YC F25

Your landing page’s weakest score is Trust.

Fix Trust

Score over time

908070605040Martian 49 on May 10 (Wayback)Martian 50 on May 21 (Wayback)Martian 48 on Jul 13 (Wayback)Martian 49 on Jul 24 (Wayback)Martian 52 on Jul 31 (Live)Not Diamond 58 on Feb 9 (Wayback)Not Diamond 58 on Feb 27 (Wayback)Not Diamond 59 on Mar 13 (Wayback)Not Diamond 58 on Apr 15 (Wayback)Not Diamond 56 on May 27 (Wayback)Not Diamond 56 on Jun 7 (Wayback)Not Diamond 62 on Jul 31 (Live)Requesty 76 on Apr 1 (Wayback)Requesty 74 on Apr 28 (Wayback)Requesty 71 on Jun 28 (Wayback)Requesty 67 on Jul 14 (Wayback)Requesty 71 on Jul 25 (Wayback)Requesty 69 on Aug 25 (Wayback)Requesty 76 on Sep 10 (Live)OpenRouter 79 on Jul 3 (Wayback)OpenRouter 79 on Jul 17 (Wayback)OpenRouter 78 on Jul 31 (Wayback)OpenRouter 76 on Aug 14 (Wayback)OpenRouter 77 on Aug 28 (Wayback)OpenRouter 78 on Sep 8 (Wayback)OpenRouter 77 on Sep 10 (Live)Unify 66 on Sep 20 (Live)dari.dev 47 on Apr 17 (Wayback)dari.dev 67 on Sep 22 (Live)Apr 17Sep 22
Past reportsShow past reportsHide past reports
  1. Sep 22, 202667LiveOverall +20 · Value props +38 · Trust +28Viewing
  2. Apr 17, 202647First tracked report — strongest CTA 58, weakest Value props 34View →

Trust

62Fair

Proof that the product is real, competent, and safe to try.

dari.dev
OpenRouter
Trust read
dari.devBenchmark chart "Next.js Self-Bench · Pi Agent" plotting accuracy vs. model cost per task (47 tasks) against Claude Opus
Martian"We've worked on Google DeepMind's early scientific initiatives, developed Anthropic's interpretability efforts, and com
Not Diamond5%+ accuracy gains | 30%+ cost savings | 2x faster dev cycles
RequestyDetailed analytics dashboard screenshot showing "Sum of Cost by Model Over Time — $445.95" with per-model usage table (o
OpenRouterStat bar: "400T+ Monthly Tokens · 10M+ Global Users · 80+ Providers · 500+ Models"
UnifyBlog index listing dated research posts: "Why we built Continual-ARC." and "The harness is not enough."

Mobile web

62Fair

The phone experience as a first-class citizen.

Mobile web annotations on the mobile capture1234
  1. Hero badge pills — Only tappable hero elements are secondary badges; no primary full-width CTA follows the headline.
  2. Clipped Claude Code tab — Tab label wraps to two lines and 'Code' is cut off at the pill's bottom edge.
  3. Truncated curl command — Install command cuts off mid-URL with no wrap or scroll cue, so the copied value is unclear.
  4. Small header Get Started — The only conversion CTA in view is a compact header button, not thumb-sized for mobile.

Information hierarchy

68Fair

Does the design help or inhibit what the page is trying to say?

Information hierarchy annotations on the desktop capture1234
  1. Dual nav CTAs — "Book A Call" and "Get Started" sit side by side at near-equal weight, splitting the primary action.
  2. Router Package pill — An outbound GitHub link placed dead-center in the hero scan path, leaking attention before any conversion.
  3. Open-Weight Router pill — Broken/missing icon glyph renders as a placeholder box, undermining polish at the hero's focal point.
  4. Benchmark tab strip — A third same-weight interactive band creeps into the fold, competing with the terminal for the next click.

CTA

72Fair

Copy, size, verb, cost-feel — does the primary CTA earn the click?

curl -fsSL https://raw.githubusercontent.com/mupt-ai/dari-cl
dari.devin-product action
Requesty
OpenRouter

Problems

  • 'Book A Call' gets primary emphasis
  • command truncates on mobile
  • tab label 'Claude Code' clips on mobile
  • no microcopy framing the install block

Value props

72Fair

Title, subtitle, hero — do they sell the product's true value?

dari.dev’s headline undersellsRun More Agents. Spend The Same.

Working

  • Outcome-first: more agents, flat spend
  • Speaks to token-cost pain directly
  • Install command shows dev-first proof

Could improve

  • No subhead explaining the router mechanism
  • Cost/accuracy numbers buried below fold
  • Assumes reader knows what a router is