GPT-6 Astra API priced at $10/$50 per million tokens; Qwen Office hits 30M users in a month
创业邦 · wechat · 2026-09-04
Chuangyebang's daily AI roundup (Sept 4):
- OpenAI: GPT‑6 Astra API pricing is higher than GPT‑5.6 Sol — $10 per million input tokens, $50 output, with cache read/write billed separately (vs $5/$30 for the previous gen).
- Alibaba: Qwen Office (enterprise agent product) surpassed 30M users in its first month, over half enterprise.
- Tencent: launched "Wujing", an autonomous AI penetration-testing product delivered as SaaS, covering pre-launch checks, continuous pentesting, attack-defense drills, and business-logic vulnerability scenarios.
- MiniMax: Saudi Arabia's HUMAIN built the first Arabic LLM, HUMAIN M3, post-trained on 1T+ Arabic tokens from MiniMax M3 — the first time a Chinese model serves as the base for an overseas sovereign model; MiniMax stock rose 10% intraday.
- Zhipu: Tmall flagship store drove 50x search growth in 48 hours and 160%+ growth in AI token subscription sales.
- China-led ISO 37127-2 smart-city AI standard approved, developed with 16 countries.
Related event: GPT-6 Astra pricing leaks: $10 input / $50 output per million tokens(7 posts)→
More from Models
- GPT-6 Astra Beats 5.6 Sol Pro (Max) on FrontierMath T4; Open Models Seen 18 Months Behind — inductionheads · 2026-09-05
- TheZvi breaks down the Claude Fable 5.1 system card: 200+ pages of safety evals — TheZvi · 2026-09-05
- COLM paper: legible chain-of-thought steps aren't necessarily important — LauraRuis · 2026-09-05
- User Impressions: Fable 5.1 on Low Reasoning Is Underrated — weswinder · 2026-09-05
- Qwen3.8-27b Is the First Local Model This User Can Blindly Trust for 8+ Hour Agent Runs — Express_Quail_1493 · 2026-09-04
- GLM runs four free-token events, unlimited coding use in daily window — Zai_org · 2026-09-04