Models

Gemini 4 Neon 浏览器使用测试达 74%,与 Claude Opus 5.5 持平但价格减半

Why we picked it

Browser Use 团队实测 Google 新模型做浏览器自动化,74% 成功率追平 Claude Opus 5.5 还便宜一半,不过 GPT-6.2 Sol 仍领先 11 个点。做网页智能体选模型前先看这份价格对比,能帮你省不少钱。

Browser Use 团队实测 Google 新发布的 Gemini 4 Neon,在浏览器自动化任务上的成功率达到 74%,与 Claude Opus 5.5 相同,但价格只有后者的一半。相比今年 9 月 Gemini 3.8 Flash 的 53%,两个月内提升了 21 个百分点,进步幅度明显。

Body

不过 Gemini 4 Neon 还不是该测试里的最好成绩:OpenAI 的 GPT-6.2 Sol 在相同价格下拿到 82%,高出 11 个百分点。Browser Use 评论指出,Google 过去在浏览器使用能力上表现较差,但每次发版都有大幅跳跃。

对做浏览器自动化智能体的开发者来说,这份实测给了清晰的价格-能力坐标系:同等预算下 GPT-6.2 Sol 能力最强,预算敏感场景可以选 Gemini 4 Neon 平替 Opus 5.5。

Original · Browser Use

Gemini 4 Neon is actually very good at browser use 👀 google used to be terrible here, but every release is a big jump > neon: 74%, same as opus 5.5 at half the price > up from 53% on 3.8 flash in september > still not sota: gpt-6.2 sol gets 82% at the same price, 11% better📊 11 ⚡

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Gemini 4 Neon, Claude Opus, GPT-6.2 Sol, Browser Use.

Source attribution: curated from Browser Use (@browser_use).