C Evidence Grade C Use with Hedging Recent

70.7%

70.7% — tasks matching or exceeding expert performance for GPT-5.2 Thinking (2025)

2025  · Unknown  · Setka

Cite this stat Share How we verify
Source: Setka View source
Curated by Aaron Agius  · Founder, Louder Online
Last verified:  ·  How we verify data →
Source verified Evidence graded
Value
70.7%
Unit
Percentage
Data Vintage
2025
Entity
GPT-5.2 Thinking
Study Type
Unknown
Trend
Stable
C
Moderate quality evidence

Useful directional data with some methodological gaps. Treat as indicative rather than definitive — corroborate before citing in formal research.

49/100 confidence
Setka Primary Tier B
https://setka.ru/posts/019b3a15-de4a-7ec2-99ee-6c61fe859c17 ↗
Confidence: 1/100 Vintage: 2025
Source Article
🔥 СЕТКА ALтушки: главные AI-движения дня Google представил Gemini 3 Deep Research Agent — улучшенного AI-агента, который сам организует и проводит исследования с точностью 46.4%. Новый Interactions API облегчит встроение этой фишки в свои приложения. 🚀 Это отличный инструмент для разработчиков, рабо…
Also cited by
硅星GenAI (Silicon Star GenAI) https://www.huxiu.com/article/4816925.html
2026-05-03
Huxiu (虎嗅) / 数字生命卡兹克 (WeChat public account) https://www.huxiu.com/article/4816662.html
2026-05-04
Original Research
OpenAI · Original study ↗

GDPval evaluation covering 44 professions across 9 core US GDP-contributing industries; human expert blind review of model-generated deliverables (sales PPTs, accounting tables, ER schedules, manufacturing charts, video content)

Study: Unknown
Suggested hedging: based on claims without stated methodology, with unverified sample sizes...
Plain Text
GPT-5.2 Thinking: 70.7% tasks matching or exceeding expert performance. Source: Setka (2025). Via Lighthouse Research Data — https://lighthousedata.io/data/gpt-5-2-thinking-gdpval-expert-parity
HTML Embed
<blockquote cite="https://lighthousedata.io/data/gpt-5-2-thinking-gdpval-expert-parity" style="border-left:3px solid #2563eb;padding:12px 16px;margin:16px 0;font-family:system-ui,sans-serif;background:#f0f4ff;"><strong>70.7%</strong> — tasks matching or exceeding expert performance for GPT-5.2 Thinking<br><small>Source: Setka (2025) · <a href="https://lighthousedata.io/data/gpt-5-2-thinking-gdpval-expert-parity" target="_blank" rel="noopener">Lighthouse Data</a></small></blockquote>
Share
Share on X Share on LinkedIn

Use this data in your work

Copy the citation, embed the widget, or explore thousands more verified marketing statistics — all free.

Copy citation Explore similar stats →