OpenAI 发布 GPT-6.2 Astra 与 Sol,登顶 Agent Arena Index 与 Code Arena Pareto 前沿
OpenAI 这次同时推了两款 GPT-6.2:Astra 冲性能榜,Sol 主打省钱,后者成本比 Astra 低七成左右还进了 Pareto 前沿。如果你在跑智能体或做 WebDev,这份基于 7.1K 真实会话的榜单数据比跑分更有参考价值,选型前值得对照看一下成本和分数的平衡点。
OpenAI 的 GPT-6.2 Astra(Max)和 GPT-6.2 Sol(Max)已登上 Arena 新推出的 Agent Arena Index。Astra 以 1291 分排第 6,成为 OpenAI 当前最佳模型,基于 7.1K+ 真实智能体任务会话评测,单任务中位成本 $2.94,较上一代 GPT-6 Astra 的 $4.91 降低 40%。Sol 以 1241 分排第 11,与 GPT-6.1 Sol 持平,但单任务成本 $0.90 比前者低 13%,进入成本 Pareto 前沿。
在 Code Arena: WebDev 榜单上,Astra 以 1781 分排第 3,比 GPT-6 Astra 高 18 分,在 Consumer Product 类目升至第 1,单任务成本 $5.45 较上代降低 42%;Sol 以 1757 分排第 7,成本 $1.45 降低 21%,两款均进入 Pareto 前沿。Astra 在 Praise vs Complaint 用户满意度信号中排第 5(1462 分)。
两款定位差异明显:Astra 比 Sol 高 50 分(Agent 榜),但 Sol 单任务成本低 69%($0.90 vs $2.94),WebDev 场景下 Sol 成本也低 73%。Agent Arena Index 基于全球社区数百万真实任务轨迹构建,是 Arena 衡量智能体综合表现的新主指标。
Scores for GPT-6.2 Astra and Sol are on the NEW Agent Arena Index! - #6 with 1291 pts: GPT-6.2 Astra (Max), now @OpenAI ’s best model - #11 with 1241 pts: GPT-6.2 Sol (Max), joining the cost Pareto frontier The Agent Arena Index is our new headline measure of overall agent performance, built from millions of real-world tasks and traces from our global community. Results for GPT-6.2 Astra and Sol are based on 7.1K+ real agentic sessions. GPT-6.2 Astra: #6 with 1291 pts - +9 pts vs previous variant GPT-6 Astra (Max), currently #7 - Median cost per task: $2.94 vs. $4.91 for GPT-6 Astra, 40% lower - Strongest signal: #5 with 1462 pts in Praise vs Complaint (user satisfaction with agent's outcome) GPT-6.2 Sol: #11 with 1241 pts - Matches #12 GPT-6.1 Sol (Max) with 1241 pts, at a 13% lower median cost per task ($0.90 vs. $1.04) - Joins the Pareto frontier (see placement in thread) 6.2 Astra vs. Sol - Astra scores +50 pts above Sol at #6 vs #11 - Sol’s median cost per task is 69% lower ($0.90 vs. $2.94) Congrats to the team at @OpenAI on this release! Arena.ai @arena Real-world Arena scores are in for GPT-6.2 Astra and GPT-6.2 Sol by @OpenAI in the Code Arena: WebDev, and both have landed on the Pareto frontier! - #3 with 1781 pts: GPT-6.2 Astra (Max) - #7 with 1757 pts: GPT-6.2 Sol (Max) GPT-6.2 Astra: #3 with 1781 pts - +18 pts vs previous variant GPT-6 Astra (Max) - #1 in Consumer Product, compared with #4 for GPT-6 Astra - Median cost per task: $5.45 vs. $9.46 for GPT-6 Astra, 42% lower - Lands on Pareto frontier (see placement in thread) GPT-6.2 Sol: #7 with 1757 pts - +13 pts vs GPT-6.1 Sol (Max) - Median cost per task: $1.45 vs. $1.84 for GPT-6.1 Sol, 21% lower - Lands on Pareto frontier (see placement in thread) 6.2 Astra vs. Sol - Astra scores +24 pts above Sol - Sol’s median cost per task is 73% lower ($1.45 vs. $5.45) Congrats to the @OpenAI team on this release. Agent score coming soon! 🔗 View Quoted Tweet📊 18 ⚡
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 GPT-6.2, OpenAI, Agent Arena Index, Code Arena.
Source attribution: curated from lmarena.ai (@lmarena_ai). Cross-source corroboration: 11 independent sources have covered this development (including Artificial Analysis, kimmonismus, TestingCatalog), confirming industry-wide signal strength.