Author: RadarAI Editorial
Editor: RadarAI Editorial
Last updated: 2026-11-19
Review status: Editorial review pending
Brief
速报
官方
AI动态
开源
Anthropic completed its IPO pricing at $83 per share, raised over $94 billion, and listed on Nasdaq, becoming the biggest capital market event of the week [3][19][22]; meanwhile, Google DeepMind's Gemini 4 Neon climbed rapidly on the Cursor and Agent Arena leaderboards, with release intervals compressed from 28 days to 6 days [1][7], while OpenAI continued to expand its ecosystem boundaries through background task discounts and safety collaboration [...
Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.
## 🔍 Core Insights
**Anthropic** completed its IPO pricing at **$83 per share**, raised over **$94 billion**, and listed on Nasdaq, becoming the biggest capital market event of the week [3][19][22]; meanwhile, **Google DeepMind**'s **Gemini 4 Neon** climbed rapidly on the Cursor and Agent Arena leaderboards, with release intervals compressed from 28 days to 6 days [1][7], while **OpenAI** continued to expand its ecosystem boundaries through background task discounts and safety collaboration [15][16][9].
## 🚀 Key Updates
- **Anthropic IPO priced at $83 per share, listing on Nasdaq tomorrow** [3]: Above the $70–78 range, with a fully diluted valuation of $1.5 trillion, ticker ANT.
- **Anthropic raises a record total of over $94 billion** [19]: Subscription orders reached up to 6 times the offered shares, with about 20% of the allocation reserved for retail investors.
- **Gemini 4 Neon launches on Cursor** [1]: Scored 63.7% on CursorBench 4.1, higher than Hydrogen's 57.3%, priced at $3/$15.
- **Gemini 4 Neon reaches No. 2 on the Agent Arena Index** [7]: The High version ranks near the top with 1,336 points, and DeepMind's release cadence has notably accelerated.
- **Google opens Gemini to 2GB-memory Android devices** [2]: Covering Android Go devices, potentially reaching hundreds of millions of users, supporting more than 45 languages.
- **OpenAI background task discount saves up to 40%** [16]: GPT-6.2 Astra and Sol cost less per token in background mode than in Standard.
- **OpenAI publishes report showing AI accelerating internal research** [8]: Researchers' use of AI has risen exponentially, with safety and alignment headcount growing fastest.
- **Palo Alto acquires Applied Compute for $4.5 billion** [24]: An all-stock deal, targeting an AI startup founded by former OpenAI researchers.
## 🔗 Sources
[1] Gemini 4 Neon launches on Cursor — https://aihot.news/items/hnez64zu373e0060mncyx0vhb
[2] Google opens Gemini app to low-memory Android devices, supporting as little as 2GB RAM — https://aihot.news/items/uul8eo6iuigzh477h51zfuxj7
[3] Anthropic sets IPO price at $83 per share, listing on Nasdaq tomorrow — https://aihot.news/items/hokjlogeo9373uw76eeb3f3mh
[7] Gemini 4 Neon reaches No. 2 on the Agent Arena Index, Google DeepMind's release cadence accelerates — https://aihot.news/items/p0r06dbw9dbo2sk23pj1e0y4y
[8] OpenAI publishes report showing how AI accelerates its internal research — https://aihot.news/items/f6lqwdjk5x80bwats1fvxagsu
[9] Greg Brockman confirms OpenAI collaborated with researchers to disclose NGINX 1.31.8 vulnerability NGINX2Shell — https://aihot.news/items/vntkimin8apqkocftgs2ipxm0
[16] OpenAI introduces Responses API background mode discount window, saving up to 40% per token on GPT-6.2 — https://aihot.news/items/o9azm5jq20btm5oez1ds8nsxe
[19
Anthropic completed its IPO pricing at $83 per share, raised over $94 billion, and listed on Nasdaq, becoming the biggest capital market event of the week [3][19][22]; meanwhile, Google DeepMind's Gemini 4 Neon climbed rapidly on the Cursor and Agent Arena leaderboards, with release intervals compressed from 28 days to 6 days [1][7], while OpenAI continued to expand its ecosystem boundaries through background task discounts and safety collaboration [15][16][9].
🚀 Key Updates
- Anthropic IPO priced at $83 per share, listing on Nasdaq tomorrow [3]: Above the $70–78 range, with a fully diluted valuation of $1.5 trillion, ticker ANT.
- Anthropic raises a record total of over $94 billion [19]: Subscription orders reached up to 6 times the offered shares, with about 20% of the allocation reserved for retail investors.
- Gemini 4 Neon launches on Cursor [1]: Scored 63.7% on CursorBench 4.1, higher than Hydrogen's 57.3%, priced at $3/$15.
- Gemini 4 Neon reaches No. 2 on the Agent Arena Index [7]: The High version ranks near the top with 1,336 points, and DeepMind's release cadence has notably accelerated.
- Google opens Gemini to 2GB-memory Android devices [2]: Covering Android Go devices, potentially reaching hundreds of millions of users, supporting more than 45 languages.
- OpenAI background task discount saves up to 40% [16]: GPT-6.2 Astra and Sol cost less per token in background mode than in Standard.
- OpenAI publishes report showing AI accelerating internal research [8]: Researchers' use of AI has risen exponentially, with safety and alignment headcount growing fastest.
- Palo Alto acquires Applied Compute for $4.5 billion [24]: An all-stock deal, targeting an AI startup founded by former OpenAI researchers.
🔗 Sources
[1] Gemini 4 Neon launches on Cursor — https://aihot.news/items/hnez64zu373e0060mncyx0vhb
[2] Google opens Gemini app to low-memory Android devices, supporting as little as 2GB RAM — https://aihot.news/items/uul8eo6iuigzh477h51zfuxj7
[3] Anthropic sets IPO price at $83 per share, listing on Nasdaq tomorrow — https://aihot.news/items/hokjlogeo9373uw76eeb3f3mh
[7] Gemini 4 Neon reaches No. 2 on the Agent Arena Index, Google DeepMind's release cadence accelerates — https://aihot.news/items/p0r06dbw9dbo2sk23pj1e0y4y
[8] OpenAI publishes report showing how AI accelerates its internal research — https://aihot.news/items/f6lqwdjk5x80bwats1fvxagsu
[9] Greg Brockman confirms OpenAI collaborated with researchers to disclose NGINX 1.31.8 vulnerability NGINX2Shell — https://aihot.news/items/vntkimin8apqkocftgs2ipxm0
[16] OpenAI introduces Responses API background mode discount window, saving up to 40% per token on GPT-6.2 — https://aihot.news/items/o9azm5jq20btm5oez1ds8nsxe
[19
← Back to Updates