DeepSeek just open-sourced its inference engine

AI summary

DeepSeek has open-sourced DeepInfer, a high-performance and large-scale inference framework designed specifically for its models.

Key points

•DeepSeek has open-sourced the inference engine DeepInfer.
•The framework is a high-performance and large-scale inference infrastructure designed for DeepSeek models.
•It is available through the deepseek-ai/DeepInfer repository on GitHub.

#LLM / Model Trends #Semiconductors & Infrastructure #DeepSeek#DeepInfer

Read the original

ITmedia AI+10/20/2026

US Department of Homeland Security orders exclusion of Chinese DeepSeek products from federal civilian agencies

The US Department of Homeland Security published an order in the Federal Register to exclude Chinese company DeepSeek's products and related services from federal civilian agencies' procurement and IT systems.

Hacker News10/20/2026

DeepSeek banned from U.S. federal civilian agencies under new DHS order

DeepSeek has been banned from use in U.S. federal civilian agencies under a new order from the Department of Homeland Security (DHS).

gihyo.jp10/19/2026

DeepSeek-V4.1-Flash-1017 with improved performance and speed arrives, DeepSeek Search API and Fetch API opened to all users

DeepSeek announced DeepSeek-V4.1-Flash-1017, an updated lightweight multimodal MoE model, and opened its Search and Fetch APIs to all users.

Hacker News10/19/2026

DeepSeek v4.1 Flash 1017 scores 45 in Artificial Analysis

An analysis compares DeepSeek V4.1 Flash 1017 (Max) with other AI models across key metrics including quality, price, performance, and context window.

ITmedia AI+10/19/2026

China's DeepSeek releases new V4.1-Flash version; model claiming to surpass some Astra capabilities is open-sourced

DeepSeek released DeepSeek-V4.1-Flash-1017, claiming to surpass GPT-6 Astra in some coding agent performance evaluations.