Nvidia and Microsoft Launch RTX Spark Chip; Anthropic Releases Claude · 1008-722

2026-10-08 16:00

Author: RadarAI Editorial Editor: RadarAI Editorial Last updated: 2026-10-08 Review status: Editorial review pending Brief 速报官方 AI动态开源

Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.

## 🔍 Core Insights **Nvidia** and **Microsoft** have teamed up to launch the RTX Spark chip and Surface Laptop Ultra, aiming to rebuild Windows PCs into local AI Agent platforms [1]; meanwhile, **Anthropic** has released **Claude Haiku 5.5** with roughly 75% lower cost, and **OpenAI** has pushed out GPT-6, escalating both the price war and capability race at the model layer [1][8]. On the application side, German search engine **Ecosia**, disappointed with Mistral, has switched to Chinese open-source models Qwen, GLM, and Kimi, cutting costs in half [4], while **Isomorphic Labs** is seeking funding at a $40-50 billion valuation, showing continued heat in the AI pharmaceutical track [10]. ## 🚀 Key Updates - **Nvidia and Microsoft jointly launch RTX Spark chip and Surface Laptop Ultra** [1]: Starting at $2,599, it focuses on local "infinite intelligence" compute power, rebuilding Windows PCs as AI Agent platforms. - **Claude Haiku 5.5 released** [8]: About 75% lower cost than Haiku 4.5, leading GPT-6 Luna across all benchmarks, billed as the cheapest and fastest small model. - **Ecosia abandons Mistral for Chinese open-source models** [4]: Due to Mistral's quality lagging a year behind competitors, it switched to Qwen, GLM, and Kimi, cutting costs in half. - **Isomorphic Labs seeks funding at $40-50 billion valuation** [10]: The Alphabet-owned AI drug discovery company is in early talks, having completed a $2.1 billion Series B in May this year. - **Stanford DeLM proposes decentralized multi-agent framework** [7]: Replacing a central coordinating agent with shared context and task queues, achieving up to 2.49x speed on complex coding tasks. - **Shengshu Technology releases Vidu Q4 Preview** [12]: Supports 3 reference audio clips and 2K/4K output, with a launch price starting at 0.09 yuan/second, roughly 5x the output for the same budget. - **Perplexity releases pplx-embed-v2-late embedding model** [22]: Includes a 0.6B edge version and a 9B version, with the 9B achieving a best score of 92.4% on MADQA. - **Ant Ling-3.1-Flash lands on OpenRouter** [6]: Released through Novita as a Day-0 partner, further expanding the domestic model ecosystem. ## 🔗 Sources [1] Jensen Huang unveils the most powerful Surface ever! Designed for "infinite intelligence" — https://www.bestblogs.dev/article/43c52545c2?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [2] Morning report | A wave of Apple new products exposed, possibly launching next week / Global pure gasoline vehicle share falls below 50% for the first time / Surface Laptop Ultra launched starting at $2,599 — https://www.bestblogs.dev/article/5c7637f298?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [3] OpenAI's Altman: The world should accept some "bad things" brought by AI in exchange for huge dividends — https://aihot.news/items/ye1vhn729at3qihhdi0u31ffy [4] German search engine Ecosia abandons Mistral, switches to Chinese open-source models Qwen, GLM, and Kimi — https://aihot.news/items/vdwwvrr8ac

Nvidia and Microsoft have teamed up to launch the RTX Spark chip and Surface Laptop Ultra, aiming to rebuild Windows PCs into local AI Agent platforms [1]; meanwhile, Anthropic has released Claude Haiku 5.5 with roughly 75% lower cost, and OpenAI has pushed out GPT-6, escalating both the price war and capability race at the model layer [1][8]. On the application side, German search engine Ecosia, disappointed with Mistral, has switched to Chinese open-source models Qwen, GLM, and Kimi, cutting costs in half [4], while Isomorphic Labs is seeking funding at a $40-50 billion valuation, showing continued heat in the AI pharmaceutical track [10].

🚀 Key Updates

Nvidia and Microsoft jointly launch RTX Spark chip and Surface Laptop Ultra [1]: Starting at $2,599, it focuses on local "infinite intelligence" compute power, rebuilding Windows PCs as AI Agent platforms.
Claude Haiku 5.5 released [8]: About 75% lower cost than Haiku 4.5, leading GPT-6 Luna across all benchmarks, billed as the cheapest and fastest small model.
Ecosia abandons Mistral for Chinese open-source models [4]: Due to Mistral's quality lagging a year behind competitors, it switched to Qwen, GLM, and Kimi, cutting costs in half.
Isomorphic Labs seeks funding at $40-50 billion valuation [10]: The Alphabet-owned AI drug discovery company is in early talks, having completed a $2.1 billion Series B in May this year.
Stanford DeLM proposes decentralized multi-agent framework [7]: Replacing a central coordinating agent with shared context and task queues, achieving up to 2.49x speed on complex coding tasks.
Shengshu Technology releases Vidu Q4 Preview [12]: Supports 3 reference audio clips and 2K/4K output, with a launch price starting at 0.09 yuan/second, roughly 5x the output for the same budget.
Perplexity releases pplx-embed-v2-late embedding model [22]: Includes a 0.6B edge version and a 9B version, with the 9B achieving a best score of 92.4% on MADQA.
Ant Ling-3.1-Flash lands on OpenRouter [6]: Released through Novita as a Day-0 partner, further expanding the domestic model ecosystem.

🔗 Sources

[1] Jensen Huang unveils the most powerful Surface ever! Designed for "infinite intelligence" — https://www.bestblogs.dev/article/43c52545c2?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [2] Morning report | A wave of Apple new products exposed, possibly launching next week / Global pure gasoline vehicle share falls below 50% for the first time / Surface Laptop Ultra launched starting at $2,599 — https://www.bestblogs.dev/article/5c7637f298?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item [3] OpenAI's Altman: The world should accept some "bad things" brought by AI in exchange for huge dividends — https://aihot.news/items/ye1vhn729at3qihhdi0u31ffy [4] German search engine Ecosia abandons Mistral, switches to Chinese open-source models Qwen, GLM, and Kimi — https://aihot.news/items/vdwwvrr8ac

← Back to Updates