AITOP WeeklyIssue 122346 stories

GPT-6.2 Astra Launches, Anthropic IPO Prices at $83, Qwen4-Max Debuts

Singapore·16 Nov – 22 Nov·Ten minutes

OpenAI's GPT-6.2 Astra swept ARC-AGI-3 with a perfect 100% on the official validation test, ranking second on the Artificial Analysis Intelligence Index behind only Claude Fable 5.2, with API input and output tokens priced 20% below GPT-6 Astra. Its system card also marks the first time OpenAI has rated a model High on self-improvement. Days later, Anthropic priced its Nasdaq IPO at $83 per share, giving regional investors a reference point for AI valuations. On the Chinese side, Alibaba's Qwen4 Max scored 43 on the Intelligence Index at $3.27 per task, while Tencent open-sourced the 770B-parameter multimodal MoE Hunyuan Hy4 with a 1M context window. For Singapore and ASEAN builders, the week delivered more flagship options and cheaper inference.

01

Five things this week

Enough to retell
1
OpenAI 发布 GPT-6.2 Astra:代码更整洁,API 输入输出降价 20%
GPT-6.2 Astra 发布,API 输入输出降价 20%,直接降低东盟团队用模型成本
@OpenAIDevs
2
GPT-6.2 Astra 在 ARC-AGI-3 验证测试拿下 100% 得分
Astra 在 ARC-AGI-3 拿下首个 100% 满分,推理能力创新纪录
ARC Prize
3
Anthropic 以每股 83 美元定价 IPO,明日登陆纳斯达克
Anthropic 以每股 83 美元 IPO,是 AI 板块估值风向标
Anthropic
4
Anthropic 数学家称与 Claude 完成黎曼猜想证明,公开约 1700 万行 Lean 代码
Anthropic 数学家与 Claude 宣称证明黎曼猜想,公开约 1700 万行 Lean 代码
IT之家
5
NVIDIA 开源 GR00T 2 机器人模型:端到端操控真实灵巧手
NVIDIA 开源 GR00T 2 机器人模型,可端到端操控真实灵巧手
Jim Fan
02

US vs China

The same week, both sides
DimensionUnited StatesChinaWhat it means for ASEAN
New modelsOpenAI ships GPT-6.2 Astra and Sol; Astra scores 100% on ARC-AGI-3 validationAlibaba launches Qwen4-Max; Tencent open-sources 770B multimodal MoE Hunyuan Hy4More flagship options and usable open-weight alternatives for self-hosting
PricesGPT-6.2 Astra API input/output tokens 20% cheaper than GPT-6 AstraQwen4 Max costs $3.27 per task, positioned as value playFalling inference costs make models affordable for ASEAN small teams
BenchmarksAstra ranks second on Intelligence Index, behind Claude Fable 5.2, five points above GPT-6 AstraQwen4 Max scores 43, up nine from Qwen3.8 Max, trailing Kimi K3.1 ProUS-China gap narrows; choose models by value, not origin
PolicyGPT-6.2 Astra system card rates self-improvement High, cybersecurity Critical—Watch safety ratings for enterprise compliance and risk assessments
03

ASEAN this week

More →
04

Policy this week

More →
05

China to ASEAN

More →
06

Builders this week

More →
2346Stories
339China
18ASEAN
267Sources
AITOP · generated by the editorial system