DeepSeek is a Chinese AI research laboratory that has developed highly capable open-source language models including DeepSeek-V3 and DeepSeek-R1, notable for their efficiency and performance.
Meta released Muse Glimmer AI on Monday and plans to open the weights for Muse Spark 1.2 in the coming weeks, as CEO Mark Zuckerberg urged Washington to ease restrictions on US open-source developers …
Meta Platforms released Muse Glimmer, a distilled version of its Muse Spark 1.2 model with 30 billion parameters, designed to run on a single graphics card for agent-like AI tasks such as schedule man…
Meta CEO Mark Zuckerberg called for lower U.S. barriers for open-source AI models to compete with Chinese rivals as the company released Muse Glimmer, a new open-weight model designed for agentic task…
Chinese humanoid robots are moving into real jobs in factories and beyond, according to a new analysis by investment bank Morgan Stanley, fueling concerns about AI's destructive potential and China ov…
ByteDance founder Zhang Yiming has banned the Seed AI team from using model distillation to train its upcoming giant model, a choice that may slow it against rivals Moonshot AI and DeepSeek but hedges…
A National Bureau of Economic Research working paper analyzing nearly 14 million Chinese patents finds that Chinese universities, not state-owned enterprises, are the primary drivers of patents in tec…
DeepSeek released V4-Flash-0731 on July 31, 2025, and the smaller model now outperforms the larger V4-Pro-Preview across all nine published agentic benchmarks, including Terminal Bench 2.1 (82.7 vs. 7…
Chinese universities account for more than a quarter of China's patents in 14 technology areas deemed critical by the Pentagon, 8 times the U.S. rate of 3.3%, according to a National Bureau of Economi…
DeepSeek's official API documentation lists V4-Flash at $0.14 per million cache-miss input tokens, $0.28 per million output tokens, and $0.0028 per million cache-hit input tokens, making it unusually …
Cache read costs now dominate agentic AI workloads, accounting for up to 81.6% of total inference spend in a 100-turn session, according to an analysis by an unnamed author. The analysis shows that fo…
Unitree Robotics priced its Shanghai IPO at 150.80 yuan a share on August 6, valuing China's leading humanoid robot maker at roughly $9 billion, with DeepSeek and state oil giant CNPC buying in as str…
DeepSeek, the Chinese AI firm known for its low-cost services, warned users on Thursday of a "significant" price increase for its AI services, without specifying the amount. The company currently char…
On July 22, White House technology official Michael Kratsios accused Chinese startup Moonshot AI of using 'industrial distillation' of Anthropic's Fable model to build its Kimi K3 model, framing the c…
An independent developer has released DeepSeek-V4-Flash-0731-Latent-Reasoning, a 284B-parameter MoE model with a 35.7M-parameter reasoning head that performs latent reasoning in a 1024-dimensional spa…
A man asked seven AI chatbots—ChatGPT, Grok, Gemini, Doubao, DeepSeek, Qianwen, and Tencent Yuanbao—for anniversary ideas on a 500 RMB ($70) budget, and each suggested unique plans ranging from a pers…
DeepSeek-V3, an open-weight large language model from Chinese AI company DeepSeek, has leaked and is matching or beating top-tier proprietary US models in coding and math benchmarks, according to a te…
DeepSeek's V4 Flash 0731 model scored 82.7% (±1.79 SE) on Terminal-Bench 2.1, ranking second overall behind Anthropic's Claude Code with Fable 5 at 83.8% (±1.16 SE), according to Ante's public benchma…
DeepMind open-sourced WeatherNext, an AI model that outperforms ECMWF's deterministic cyclone forecasts while running orders of magnitude cheaper, though it lacks ensemble uncertainty handling for 10+…
Chinese AI labs are ending their era of low-cost models, with Moonshot's Kimi K3 priced at $3 and $14 per million input and output tokens and requiring a 30% revenue share for inference, while DeepSee…
Dwarkesh Patel predicts that continual learning, where AI systems update their weights from real-world sessions, will render current safety regulations obsolete, force alignment research to handle con…