ENLatest AI NewsRecord 10 Billion Downloads: MIIT Releases OpenHarmony Mid-Year Report, Devices Exceed 1.35 BillionAt a State Council Information Office briefing, the MIIT released H1 2026 data, providing the first comprehensive quantification of China’s open-source ecosystem. It highlights coordinated progress in basic software, industrial software, and open-source development, offering a clear basis for external assessment…..6 hours ago59.0KFirst Packet Delay 300ms, Supports 20 Dialects: Tongyi Qianwen Qwen-Audio-3.0-TTS Officially OpenedAlibaba’s Tongyi Qianwen launched Qwen-Audio-3.0-TTS, a real-time TTS model advancing from speech to expression. Its Plus version tops Artificial Analysis Speech Arena, beating Gemini 3.1 TTS. Flash version offers low-latency (~300ms first packet) interaction; Plus focuses on high naturalness and timbre fidelity…..7 hours ago12.0KA Line of GitHub Code Reveals AMD: Anthropic Exposed as Its New Customer, Accelerating the Shift Away from NVIDIAGitHub code from an AMD AI executive accidentally revealed Anthropic as a customer, signaling its effort to diversify chip supply beyond a single vendor, per SemiAnalysis…..8 hours ago14.2KTencent WorkBuddy’s Monthly Visits Exceed 20 Million in June, Leading the AI Office Intelligent Agent MarketAccording to the “Q2 2026 China Office Intelligent Agent Market Insight Report”, Tencent WorkBuddy leads the PC-based AI-native office intelligent agent market. Its monthly visits reached 20.97 million in June, surpassing the combined total of the second and third places. The total market visits exceeded 60 million in the same month, indicating a continued expansion of the industry scale. CodeBuddy, Tencent’s AI programming intelligent agent, also ranks among the top. WorkBuddy can complete office tasks through natural language commands.9 hours ago15.4KBreaking the 15-Second Magic Spell: Zhixiang Future Launches the World’s First Unlimited-Length Creative AI Agent vivago R1, Commercial Availability Increased to 85%At WAIC 2026, Zhixiang Future unveiled Vivago R1, the world’s first multimodal agent for unlimited-duration content creation. It co-founded the Belt & Road Token Alliance with BrainCore and initiated the Physical Intelligence Innovation Consortium with FJKS, advancing AI from virtual to physical through tech innovation and ecosystem synergy…..9 hours ago11.5KHugging Face Discloses AI Agent Attack Incident, Uses GLM5.2 for Log Forensic AnalysisAfter a hacker AI attack, Hugging Face tried using a US commercial large model API to analyze 17,000 logs but failed due to security mechanisms misjudging attackers vs. defenders. It then deployed the Chinese open-source model GLM5.2 on its own infrastructure, successfully completing massive log forensics. The incident highlights performance differences between AI models in security scenarios…..9 hours ago12.3K88 million dollars with open source: Ollama enters 85% of the Fortune 500, calling for everyone to get on boardOllama enables local execution of open-source LLMs with a single command, freeing developers from API keys and high costs. On July 9, it raised $88M led by Benchmark’s Peter Fenton, Theory Ventures’ Tomasz Tunguz, 8VC’s Alex Kolicich, with participation from Docker founder Solomon Hykes, ClickHouse CEO Aaron Katz, and GIMP co-creator…..10 hours ago11.9KDeepSeek V4 Final Version Tested Exposes, Possibly to Be Released Next Monday Targeting Kimi K3DeepSeek V4 test video leaked, likely launching next Monday, riding Kimi K3’s buzz. Benchmark performance rivals Claude Opus 4.8, stunning core metrics draw industry attention…..10 hours ago15.8KByteDance Launches Seed Audio 1.0: From Speaking to Creating in Audio GenerationByteDance has launched the audio creation model Seed Audio 1.0, which is now open for testing on Volcano Fangzhou. It breaks the traditional process of generating voice, sound effects, and ambient sounds with separate models and manually assembling them, achieving integrated creation of sound scenes, ensuring narrative consistency, and marking a new stage in AI audio generation with complete sound scenes.10 hours ago10.7KUsers who shouted ‘One more two weeks’ won: Tencent HuanYuan Hy3 is free for an extended period until August 5th, the 295B MoE model’s free trial period is renewedTencent’s Hunyuan Hy3 model was open-sourced with a two-week free trial. Users demanded an extension, and PR director Zhang Jun said feedback prompted a delay. Free access for WorkBuddy and CodeBuddy is now extended to Aug 5, making it nearly a month free…..12 hours ago20.1KMusk Says 2 Trillion Parameter Large Model is About to Complete Training, May Challenge Kimi K3 PerformanceAfter Moonshot AI released open-source Kimi K3 (2.8T params), Musk disclosed his 2T-param model nears completion, potentially surpassing K3 and prior 1.5T version. He called K3 “impressive,” fueling the ultra-large model race…..12 hours ago13.1KStop Counting Tokens: OpenAI Releases the Score