Flash

The news, the context, and what changes.

RSS
Flash

Amap Launches AI App Generation Platform 'Daima'

Amap is internally testing a Vibe Coding product named 'Daima'. The platform focuses on zero-threshold app building driven by natural language, targeting WeChat Mini Programs and iOS native scenarios. It enables users to quickly generate applications that are ready for launch and functional on real devices.

88Z AI editorial1 min read
Flash

Alibaba Tongyi Qianwen Launches Standalone AI Voice Input Method 'Qianwen Input'

Alibaba's Tongyi Qianwen team has launched a standalone AI voice input method called 'Qianwen Input', emphasizing casual speech and polished writing. Built on the Qianwen large model's CosyVoice capabilities, it converts speech to text while understanding semantics, removing filler words, and correcting slips to structure spoken language into written text. It supports speeds up to 300 characters per minute, recognition of nine dialects, and mixed Chinese-English speech.

88Z AI editorial1 min read
Flash

Shengshu Technology Launches AI Companion App Vidy

Shengshu Technology has launched Vidy, an AI companion app positioned as an 'AI Vibe Buddy' that supports text, voice, and real-time calls. Featuring characters like Humi and a relationship growth memory system, it leverages the self-developed U-ViT architecture and Vidu video large model to plan for real-time AI facial expression and motion generation, upgrading companionship to face-to-face video interactions.

88Z AI editorial1 min read
Flash

X Variable Completes Four Consecutive Funding Rounds, Valuation Exceeds 20 Billion Yuan

X Variable completed four consecutive funding rounds within two months, with its post-money valuation exceeding 20 billion yuan, making it the first embodied brain company in the Greater Bay Area to reach this milestone. Meituan, Alibaba, ByteDance, and Xiaomi each led a round, with Xiaomi investing in three consecutive rounds. Over thirty institutions, including 58 Group, Chery, China Mobile, and Sequoia China, participated.

88Z AI editorial1 min read
Flash

DeepReinforce Open-Sources Agentic Programming Model Series Ornith-1.0

DeepReinforce has released the open-source Agentic programming model Ornith-1.0, built on Gemma 4 and Qwen 3.5 bases with four sizes ranging from 9B to 397B. The series performs strongly on Terminal-Bench 2.1 and SWE-Bench Verified benchmarks, with the flagship 397B version surpassing Claude Opus 4.7 and the 35B MoE version outperforming Qwen 3.5-397B.

88Z AI editorial1 min read
Flash

Tencent Hunyuan Open-Sources 4B-Parameter Mobile Agent Model PhoneBuddy

Tencent Hunyuan has officially open-sourced PhoneBuddy, a mobile agent model with 4 billion parameters. Alongside the release, the team published five academic papers, including PhoneWorld and PhoneHarness, establishing a comprehensive technical framework covering environment construction, training, execution, and privacy security. The model's core innovation lies in its Real+Mock hybrid reinforcement learning strategy, which combines operations in real apps with a self-built PhoneWorld simulati

88Z AI editorial1 min read
Flash

Mistral AI Launches New Document Recognition Model Mistral OCR 4

Mistral AI has released Mistral OCR 4, a new document recognition model supporting 170 languages. It scored 93.07 on the OmniDocBench benchmark, outperforming some competitors in human preference. The model provides bounding boxes, region classification, and confidence scores, making it suitable for downstream tasks like RAG.

88Z AI editorial1 min read
Flash

OpenAI Launches First In-House Inference Acceleration Chip Jalapeño

OpenAI has unveiled Jalapeño, its first in-house inference chip, developed with Broadcom and Celestica under the leadership of a former Google TPU core member. Tailored for models like ChatGPT, it set a record for fastest ASIC tape-out and is scheduled for deployment by late 2026 with superior energy efficiency.

88Z AI editorial1 min read
Flash

Huawei Announces Full Model Availability of HarmonyOS Lobster 'Xiaoyi Claw'

Huawei has announced that the HarmonyOS 'Lobster' Xiaoyi Claw is now available for all models. Devices running HarmonyOS 5.0 or later can access it by updating the Xiaoyi App to the specified version. The updated service packages include a 49 yuan trial pack with Auto-Model mode and a 199 yuan standard pack offering choice among four models, including openPangu-2.0-Pro.

88Z AI editorial1 min read
Flash

Alibaba Cloud Launches AI Voice Input Method CosyVoice

Alibaba Cloud has released CosyVoice, a new AI voice input method powered by the Qwen large model. It offers semantic-level understanding and text regeneration, supporting real-time transcription, redundancy filtering, and multi-dialect recognition to surpass traditional acoustic transcription.

88Z AI editorial1 min read
Flash

Alibaba Cloud Launches AI Interactive Story Creation and Reading Platform 'Xiangwei'

Alibaba Cloud has launched 'Xiangwei', an AI-powered platform for interactive story creation and reading. It supports three modes: starting from scratch, continuing existing text, and planning chapters for long novels, catering to web novels, scripts, and children's stories. During reading, AI generates background images in real time, allowing readers to choose plot directions or input custom text at key points, with the platform instantly continuing the story while maintaining character consist

88Z AI editorial1 min read
Flash

NetEase Youdao Open-Sources TTS Speech Synthesis Engine Confucius4-TTS

NetEase Youdao has officially open-sourced the Confucius4-TTS speech synthesis model. The model achieves breakthroughs in zero-shot cloning, multi-language accent-free synthesis, and emotion transfer, utilizing an end-to-end architecture that supports local offline deployment.

88Z AI editorial1 min read