deepseek - Tech News from Simmons Systems

Tech News from Simmons Systems

deepseek

A collection of 49 posts

Anthropic Reports Record Distillation Campaigns by Chinese AI Labs

Anthropic releases a report Thursday alleging persistent and escalating distillation attacks by China-based AI companies aimed at harvesting the capabilities of its Claude models. The company attributes the activity to five separate campaigns totaling nearly 200 million exchanges, making the efforts larger and more aggressive than earlier incidents it describes

OpenAI and Anthropic Slash Prices as Chinese AI Models Win Customers

OpenAI and Anthropic are slashing prices on their flagship AI models as cost-conscious customers flee to cheaper alternatives from Chinese developers such as Moonshot and DeepSeek. OpenAI cuts prices for GPT-5.6 Luna, its "fastest and most affordable model," by 80 percent, while Anthropic launches Claude Opus 5,

DeepSeek Founder Admits Compute Scarcity Drives Efficiency Strategy

DeepSeek founder Liang Wenfeng reveals in a newly surfaced investor transcript that the company's celebrated efficiency is born from necessity, not strategy. In a nearly four-hour meeting covering AGI plans, chip supply, and the company's massive $7.4 billion funding round, Liang states plainly that all

DeepSeek V3.1-Terminus Update Improves Tool Use, Fixes Language Errors

Chinese AI startup DeepSeek releases V3.1-Terminus, an upgrade to its open-source model that enhances agentic coding and search tasks while fixing a bug that inserted Chinese words into English text. The update is available immediately across multiple platforms.

Mistral Launches Le Chat Mobile App to Challenge OpenAI and DeepSeek

French AI startup Mistral releases a free Le Chat mobile app for iOS and Android alongside new Pro and Enterprise tiers. The launch positions Mistral as a secure, privacy-focused alternative to ChatGPT and DeepSeek for business users.

Nvidia Suffers Record $589 Billion Loss as DeepSeek Challenges AI Spending

Nvidia experiences the largest single-day market cap loss in history as a breakthrough AI model from China's DeepSeek raises concerns about the need for expensive chips. The news triggers a massive sell-off across the semiconductor sector.

DeepSeek Expands Context Window to One Million Tokens Amid V4 Development

DeepSeek updates its web and mobile platforms with a massive 1M token context window, an eightfold increase that allows users to process entire books and codebases at once. Meanwhile, the company continues developing its highly anticipated trillion-parameter V4 flagship model.

DeepSeek Expands AI Memory Tenfold Amid Zhipu AI's GLM-5 Launch

DeepSeek increases its AI model's context window to over 1 million tokens, enabling much more complex reasoning. Rival Zhipu AI simultaneously releases its upgraded GLM-5 model to keep pace in the competitive Chinese AI market.

DeepSeek Unveils V3.2 Models With Record-Breaking Math Reasoning

DeepSeek releases two new AI models that offer advanced reasoning and cost efficiency, alongside a math model that outperforms top human competitors.

DeepSeek Launches V3.2-Exp Model With Sparse Attention, Slashes API Prices

DeepSeek releases its new V3.2-Exp experimental model featuring a novel sparse attention mechanism that cuts compute costs without sacrificing quality. The update also brings a massive 50 percent reduction in API pricing.

DeepSeek Launches V3.1 Model with Hybrid Think and Non-Think Modes

DeepSeek releases V3.1, introducing a hybrid inference system that lets users toggle between thinking and non-thinking modes. The update brings stronger agent capabilities, Anthropic API support, and new open-source weights.

Qwen3 Launches Open-Weight AI Models Rivaling Top Industry Competitors

The Qwen team releases Qwen3, a new family of open-weight large language models that achieve top-tier benchmark results in coding and math. The release includes efficient Mixture-of-Experts models and smaller dense models available under the Apache 2.0 license.

DeepSeek Unveils V3-0324 Update with Enhanced Reasoning and MIT License

DeepSeek releases the V3-0324 model update, bringing major improvements to reasoning, front-end development, and tool-use capabilities. The new model weights are now available under the permissive MIT License.

Baidu Launches Ernie 4.5 and X1 Models, Makes Chatbot Free

Chinese tech giant Baidu introduces two new AI models, Ernie 4.5 and Ernie X1, and opens its Ernie Bot to the public at no cost ahead of schedule. The new models outperform rivals in several benchmarks while remaining highly affordable.

Paris AI Summit Shifts Focus From Safety Risks to Global Opportunity

World leaders at the Paris AI Action Summit prioritize innovation and national interests over AI safety concerns, marking a major shift in global technology policy. Stanford HAI experts note that open-source advancements and geopolitical competition drive this changing landscape.