Apple Inc. is building a proprietary large language model for China in partnership with Alibaba Group, a strategic pivot from white-labeling a local model to gain control over user experience and data…
Alibaba's Qwen project scheduled the release of Qwen3.8-27B, a dense 27-billion-parameter vision-language model, for August 14, 2026, according to a countdown post from the official Qwen account. The …
Apple has built its own proprietary AI model for China with support from Alibaba, becoming the first foreign company approved to offer such a model in the country, according to Reuters. The model, dev…
Apple has trained its own large language model for the China market with help from Alibaba, making it the first foreign company approved by the Chinese government to offer a proprietary AI model in th…
Apple is building a custom large language model for China using Alibaba's cloud infrastructure and Qwen AI, according to a report. The move allows Apple to maintain privacy and regulatory compliance w…
Apple has trained a large language model specifically for the China market in partnership with Alibaba Group, three sources said, a departure from its previous reliance on third-party models. The move…
Researchers propose RA-DPO (Reliability-Aware Direct Preference Optimization), which integrates annotator agreement, model confidence, and token-level uncertainty to improve sexism detection. On 6,920…
Researchers introduced WidgetGen, a lightweight tool-grounded framework that extracts observable text and color evidence, performs high-level layout and optional chart reasoning, and directly generate…
Alibaba's Qwen team released the open weights for Qwen3.8-2.4T-A95B, a 2.4T-parameter MoE flagship, on August 13, 2026, and Fixstars deployed it on a single 8x NVIDIA B300 node, achieving twice the th…
A developer overhauled their hybrid retrieval pipeline for technical documentation, combining Qwen embeddings, BM25 sparse vectors, and reciprocal rank fusion with a FastEmbed reranker. The upgrade im…
DeepSeek announced on August 6 that it will raise API prices for its V4 model lineup, with the V4-Pro output token price during peak hours climbing to $3.96 per million tokens from $0.87, a 4.5x incre…
A developer built a model catalog drift monitor for OpenAI-compatible Chinese AI APIs, designed to catch changes in model names, context windows, output limits, and pricing before they cause incidents…
Mozilla's chief technology officer, Raffi Krikorian, argues that AI should be built like the internet, with open-source models as a viable alternative to proprietary systems. A Mozilla report publishe…
BrowserMesh, an open-source local browser runtime for external AI clients, lets MCP-compatible tools such as Claude Code, Codex, Cursor, and Qwen control multiple isolated Chromium sessions via explic…
SiliconFlow, an AI inference platform serving more than 200 open language and multimodal models through one OpenAI-compatible API, raised over 2 billion yuan (about $294 million) in a Series B in June…
Alibaba's Qwen team released Qwen3.8-2.4T-A95B, an open-weight mixture-of-experts model with 2.4 trillion total parameters and 95 billion activated per token, which scores 93.0 on PaperBench and 86.6 …
Alibaba's Qwen team launched Qwen3.8-Max on August 3, a 2.4 trillion parameter sparse Mixture-of-Experts model that runs at over 4,000 tokens per second per GPU on Nvidia's GB300 NVL72 hardware, scori…
Alibaba released the open-source Qwen3.8-2.4T-A95B model weights on August 12, a 2.4-trillion-parameter sparse mixture-of-experts model with about 95 billion active parameters per token, making it the…
Amazon Web Services (AWS) introduced a tiered KV cache architecture on Amazon SageMaker HyperPod that extends the cache hierarchy beyond GPU and CPU memory into a shared, distributed NVMe pool using C…
A developer's guide to self-hosting LLMs on budget VPS plans in 2026 finds that quantized 3B and 7B models are now viable on servers under $6 per month, with RAM bandwidth as the key bottleneck. The a…