Gemini 4 Argon 数据分析实测：称幻觉率 15%、Vals Index 第一

Stats Wire @StatsWire原推发布于 2026.10.15

在 X 上打开原推

I’ve been testing Gemini 4 Argon on heavy data analysis, and it’s easily Google’s strongest enterprise model yet. Why it beats GPT-6 Astra and Claude for serious data work: 1️⃣ 15% Hallucination Rate → 3.4× lower than Astra (51%) and Sol (54%) on AA-Omniscience 2️⃣ #1 on Vals Index (68.9%) & #1 on Finance Agent (65.4%) 3️⃣ 1M Context / 64K Output → Ingests massive CSVs and multi-file reports in 1 pass without dropping rows 4️⃣ Native Workspace Memory → Connects directly across Google Sheets, Drive & Docs at $1.99/task Coding gets all the hype, but data analysis is where Argon quietly took the crown.