Qwen3.5: Alibaba's agent leap for business in China
anjin.digitalAnalysis of Alibaba's Qwen3.5 release, which adds built-in agentic capabilities to shift from reactive chatbots to autonomous multi-step service agents, embedding Alibaba's stack across industries in China.
Alibaba's Qwen3.5 targets enterprise agent workflows with expanded multimodal support
computerworld.comAlibaba unveiled Qwen3.5, a multimodal AI model intended as a foundation for digital agents capable of advanced reasoning and tool use, with expanded multilingual support and a hosted version for enterprise developers.
Alibaba's Qwen3.5 targets enterprise agent workflows with expanded multimodal support
infoworld.comAlibaba unveiled Qwen3.5 as a foundation for digital agents with advanced reasoning and tool use, expanding multilingual support from 119 to 201 languages. The open-weight model is for developers while a hosted version targets enterprise multi-step applications.
Alibaba unveils Qwen3.5 as China's chatbot race shifts to AI agents
cnbc.comAlibaba launched Qwen3.5, a 397 billion parameter open-weight model with native multimodal and agentic capabilities, plus a hosted Qwen-3.5-Plus via Model Studio. The model claims performance on par with leading OpenAI, Anthropic and Google DeepMind models.
Chinese AI startup Zhipu releases new flagship model GLM-5
reuters.comReuters reports Zhipu AI released the open-source GLM-5 model with enhanced coding and long-running agent capabilities, approaching Anthropic's Claude Opus 4.5 on coding benchmarks and surpassing Google's Gemini 3 Pro on some tests, trained on domestically made chips.
Llama 4: Did Meta just push the panic button?
interconnects.aiAn analysis of Meta's Llama 4 launch, arguing the release feels lost amid extreme competition from open-weight models like Qwen 3, with weak released evaluations for its flagship 10M-token context window and EU restrictions on vision capabilities.
DeepSeek V4.1 Flash Specs: KV Cache Compression Explained | MindStudio
mindstudio.aiExplains DeepSeek V4.1 Flash's architecture, cutting its global KV cache footprint to roughly 890 bytes per token, about a quarter of its predecessor DeepSeek V4 Flash.
deepseek-ai/DeepSeek-V4.1-Flash · Hugging Face
huggingface.coModel card for DeepSeek-V4.1-Flash detailing its 40-layer Causal Encoder-Decoder architecture, CSA2 sparse attention, and memory-efficient KV cache designs.
DeepSeek V4.1 Flash (max) - Intelligence, Performance & Price Analysis | Artificial Analysis
artificialanalysis.aiAnalysis of DeepSeek V4.1 Flash scoring 39 on the Artificial Analysis Intelligence Index, with 552B parameters (16B active), open weights under MIT license, and a 1M-token context window.
DeepSeek V4.1 Flash - API Pricing & Benchmarks | OpenRouter
openrouter.aiOpenRouter endpoint page for DeepSeek V4.1 Flash, the first model built on DeepSeek's Causal Encoder-Decoder architecture with 552B parameters (8B active), native image understanding, and a 1M-token context window.
Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.
deepseek.comDeepSeek's official announcement of DeepSeek-V4.1-Flash, the smallest model in its new architecture family, featuring native visual understanding, faster inference, and higher throughput.
Grok 4.7, Explained in 5 Points
basenor.comBreaks down Grok 4.7's larger 2.1 trillion-parameter base model, expanded 500,000-token context, four reasoning levels, native Grok Bot harness training, and competitive API pricing tiers.
Grok 4.7 Is Out: Features, Pricing & How to Try It
kingy.aiCovers Grok 4.7's launch on September 21, 2026, with access through Cursor, Grok Build, and the xAI API, plus standard API pricing starting at $2/$6 per million tokens.
Grok 4.7: AA-Briefcase 1,657, Half Opus 5's Cost per Task
orcarouter.aiBenchmarks Grok 4.7 against peers: it scored 46 on Artificial Analysis' Intelligence Index, 56 on the Coding Agent Index, and 1,657 Elo on AA-Briefcase, at the same pricing and speed as Grok 4.6.
Introducing Grok 4.7
x.aixAI's official announcement of Grok 4.7, its most powerful model for coding and knowledge work, built on a larger base model with a longer reinforcement-learning run, improved self-verification, and a new safeguard stack.
Grok 4.7 - API Pricing & Providers | OpenRouter
openrouter.aiOpenRouter endpoint page for xAI's Grok 4.7 flagship model, describing its strength in long-running software engineering, self-verification, and long-context knowledge work, with pricing and availability details.
Grok 4.7 Is Here: What Changed and What to Know
basenor.comReviews xAI's Grok 4.7, released September 21, 2026, positioning it as the most capable model for coding, agentic tasks, and knowledge work with a 500,000-token context window and native Grokipedia 0.3.
Stability AI latest funding backed by entertainment industry biggest names
stability.aiStability AI announced a $76M Series B fundraise bringing total funding to $232M under CEO Prem Akkaraju, to fuel its generative AI product suite, applied research and professional services across music, gaming and entertainment.
Midjourney Seeks to Reveal Studios' Use of AI in Copyright Battle
variety.comMidjourney is seeking court disclosure of how Disney, Universal and Warner Bros. use AI as it defends against a major copyright lawsuit, arguing the studios engage in the same AI practices Midjourney claims are fair use.
AWS rolls out Nova Premier on Amazon Bedrock
constellationr.comAWS has rolled out Amazon Nova Premier, one of its most capable multimodal foundation models, on Amazon Bedrock, alongside other AI industry news including MongoDB's Atlas Agent Engine and AMD's acquisition of World Labs.