PulseAugur
EN
LIVE 10:03:59

SemiAnalysis launches GDPx benchmark, revealing low scores for top AI agents

By PulseAugur Editorial · [2 sources] ·

SemiAnalysis has launched GDPx, a new benchmark designed to evaluate AI agents on real-world professional tasks across 11 domains, including chip design and quantitative trading. The benchmark measures performance over extended periods, with tasks having an 8-hour budget. Claude Opus 5.5 currently leads the GDPx leaderboard with a 12% score, followed by Gemini 4 Carbon and GPT-6 Astra, highlighting the significant room for improvement in current frontier models for complex, long-horizon work. The benchmark also reveals cost-effectiveness, showing substantial differences in cost per task among top models, which is crucial for agentic deployments. AI

IMPACT Provides a new, cost-aware evaluation metric for AI agents, impacting model selection for professional workflows and highlighting current limitations in domains like finance and law.

RANK_REASON New benchmark release for evaluating AI models on professional tasks.

Read on Mastodon — mastodon.social →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

SemiAnalysis launches GDPx benchmark, revealing low scores for top AI agents

How we ranked this

Signal score
2 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
New benchmark release for evaluating AI models on professional tasks.
Source corroboration
2 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
product, other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.

Full methodology in our editorial standards.

COVERAGE [2]

  1. dev.to — LLM tag TIER_1 English(EN) · News ·

    SemiAnalysis launches GDPx, an agentic benchmark across 11 professional domains

    <p><a class="article-body-image-wrapper" href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fcdn.bsky.app%2Fimg%2Ffeed_fullsize%2Fplain%2Fdid%3Aplc%3Azl7pjaqajbk5mlwalj62hiz2%2Fbafkreihb3pwjl5tmib2sk3xcq5sz2323…

  2. Mastodon — mastodon.social TIER_1 English(EN) · Tokenmaxradar ·

    SemiAnalysis's new GDPx benchmark tests AI agents across 11 domains, from chip design to cybersecurity. Opus 5.5 tops it at just 12%. How it works: benchmarks a

    SemiAnalysis's new GDPx benchmark tests AI agents across 11 domains, from chip design to cybersecurity. Opus 5.5 tops it at just 12%. How it works: benchmarks and leaderboards. # Benchmarks # AIAgents # Semiconductors # Cybersecurity # AI https:// tokenmaxradar.com Sources (cover…

    LINKS tokenmaxradar.com newsletter.semianalysis.com/…/gdpx-measur…

RELATED ENTITIES

RELATED TOPICS