import { useState, useEffect, useRef } from "react"; import { Helmet } from "react-helmet-async"; import { Link } from "react-router-dom"; import { useTranslation } from "react-i18next"; import { useMarketingCta } from "../../hooks/useMarketingCta"; import { ArrowRight, ChevronRight, Trophy, ExternalLink, Github, BookOpen, Lightbulb, ListChecks, BookText, Target, Layers, Presentation, Quote, Cpu, Brain, } from "lucide-react"; // --------------------------------------------------------------------------- // Animated leaderboard bar chart // Uses IntersectionObserver to trigger bars growing from 0 → target width // --------------------------------------------------------------------------- interface LeaderboardEntry { name: string; score: number; /** Whether this is the Grep entry (highlighted) */ isGrep?: boolean; } function AnimatedLeaderboard({ data, /** The right-hand label format — "percent" renders 78.6%, "number" renders 56.27 */ format = "percent", }: { data: LeaderboardEntry[]; format?: "percent" | "number"; }) { const ref = useRef(null); const [animated, setAnimated] = useState(false); useEffect(() => { const el = ref.current; if (!el) return; const observer = new IntersectionObserver( ([entry]) => { if (entry.isIntersecting) { setAnimated(true); observer.disconnect(); } }, { threshold: 0.3 }, ); observer.observe(el); return () => observer.disconnect(); }, []); const maxScore = Math.max(...data.map((d) => d.score)); return (
{data.map((item, i) => { const barWidth = (item.score / maxScore) * 100; const label = format === "percent" ? `${animated ? item.score : 0}%` : animated ? item.score.toFixed(2) : "0.00"; return (
{/* Animated bar background */}
{item.name} {item.isGrep && }
{label}
); })}
); } // --------------------------------------------------------------------------- // Metric card (used in the 4-card grids) // --------------------------------------------------------------------------- function MetricCard({ icon: Icon, title, value, delta, }: { icon: React.ComponentType<{ className?: string }>; title: string; value: string; delta?: string; }) { return (

{title}

{value}
{delta && (
{delta}
)}
); } // --------------------------------------------------------------------------- // Hero stat card (the three score pills at the top) // --------------------------------------------------------------------------- function HeroStatCard({ benchmark, score, unit, }: { benchmark: string; score: string; unit?: string; }) { return (
{score} {unit && {unit}}
{benchmark}
); } // =========================================================================== // BENCHMARK PAGE // =========================================================================== export default function BenchmarkPage() { const { t } = useTranslation("marketing"); const { href: ctaHref, state: ctaState } = useMarketingCta(); const ctaText = ctaState === "signedIn" ? t("benchmark.cta.getStarted", "Get Started") : ctaState === "returning" ? t("common:header.signIn", "Sign In") : t("benchmark.cta.joinWaitlist", "Get Started"); // -- DRACO data (top 5) -- const dracoData: LeaderboardEntry[] = [ { name: "Grep", score: 78.6, isGrep: true }, { name: "Perplexity DR (Opus 4.6)", score: 70.5 }, { name: "Claude Opus 4.6", score: 59.8 }, { name: "Gemini Deep Research", score: 59.0 }, { name: "OpenAI Deep Research (o3)", score: 52.1 }, ]; // -- DeepSearchQA data (top 5) -- const deepSearchData: LeaderboardEntry[] = [ { name: "Grep", score: 84.5, isGrep: true }, { name: "Perplexity Deep Research", score: 81.9 }, { name: "Moonshot K2.5", score: 77.1 }, { name: "Anthropic Opus 4.5", score: 76.1 }, { name: "Parallel Ultra2x", score: 72.6 }, ]; // -- DeepResearch Bench (RACE) data (top 5) -- const raceData: LeaderboardEntry[] = [ { name: "Grep", score: 56.27, isGrep: true }, { name: "Cellcog Max", score: 56.13 }, { name: "nvidia-aiq", score: 55.95 }, { name: "Cellcog", score: 55.31 }, { name: "CMCC-DeepInsight", score: 55.24 }, ]; return ( <> #1 on Every Major Benchmark — DRACO, DeepSearchQA, DeepResearch Bench | Grep AI
{/* ================================================================ 1. HERO — dark background ================================================================= */}
{/* Breadcrumb */}
Home Benchmark

#1 on Every Major Benchmark

Evaluated on DRACO, DeepSearchQA, and DeepResearch Bench — PhD-level research tasks graded by domain experts.

{/* Three headline stat cards */}
{/* CTAs */}
{/* ================================================================ 2. DRACO BENCHMARK ================================================================= */}
{/* Badge */}
DRACO — Perplexity + Harvard

DRACO Benchmark

100 open-ended research questions across 10 domains, judged by Gemini-2.5-Pro against 3,934 weighted rubric criteria. Grep leads all four evaluation axes and wins 9 of 10 domains.

Grep wins 9 of 10 domains

{/* Leaderboard */}
{/* Performance by evaluation axis */}
{/* ================================================================ 3. DEEPSEARCHQA BENCHMARK ================================================================= */}
{/* Badge */}
DeepSearchQA — Google
{/* Leaderboard first on md+ for visual variety */}

DeepSearchQA

896 multi-step research questions across 17 subject domains. Judge: Gemini 2.5 Flash. Grep achieves 84.5% FC with perfect scores in Linguistics, Biology, and Arts & Entertainment.

14 of 17 categories exceed 80% FC

{/* ================================================================ 4. DEEPRESEARCH BENCH (RACE) ================================================================= */}
{/* Badge */}
DeepResearch Bench — RACE Framework

DeepResearch Bench

100 PhD-level research questions (50 Chinese, 50 English), judged by Gemini-2.5-Pro. A score above 50 means the system outperformed the human expert. Grep leads the field of 34 systems.

{/* Leaderboard */}
{/* RACE dimensions — 4 cards */}

Insight

58.98

Comprehensiveness

56.79

Instruction Following

53.49

Readability

53.50
{/* ================================================================ 5. METHODOLOGY ================================================================= */}

Methodology

Multi-Agent Architecture

Grep orchestrates specialised sub-agents — each responsible for search, synthesis, verification, and citation — then merges their outputs into a single, coherent research report.

Claude Opus 4.6 Backbone

All reasoning and synthesis steps are powered by Claude Opus 4.6, giving Grep best-in-class analytical depth, nuanced judgement, and instruction following.

{/* ================================================================ 6. CTA ================================================================= */}

Experience #1 Ranked Research

See why Grep outperforms OpenAI, Google, Perplexity, and every specialised research platform on PhD-level tasks.

{ctaText} API Docs
); }