DeepSeek Open-Sources DeepInfer for Nvidia and Huawei AI Chips
DeepSeek has released DeepInfer, its production inference engine, with support for V4.1 Flash on Nvidia SM100 and Huawei Ascend 950 hardware. The public release covers offline inference.
DeepSeek
DeepSeek is a Chinese AI lab known for efficient large language models, reasoning models, and open-weight AI releases.
- Founded 2023
- Headquarters Hangzhou, Zhejiang, China
- CEO Liang Wenfeng
Overview
-
Founded
2023
-
Headquarters
Hangzhou, Zhejiang, China
-
Industry
Artificial Intelligence
-
CEO
Liang Wenfeng
-
Founders
Liang Wenfeng
-
Funding
Backed by High-Flyer-related resources
-
Valuation
N/A
-
Employees
N/A
About DeepSeek
DeepSeek is a Chinese AI lab known for efficient large language models, reasoning models, and open-weight releases that have drawn global attention from developers, researchers, and enterprise AI teams. The company was founded by Liang Wenfeng and is associated with High-Flyer, combining frontier AI research with a strong focus on cost-efficient model training and deployment.
TrueUp lists DeepSeek as a generative AI company with the business focus “AI lab (part of High-Flyer Capital)” and valuation N/A in its visible generative AI company ranking.
Click to see more
Products & Business
-
Products & Services
- DeepSeek models
- reasoning models
- chat assistant
- model APIs
-
Platform & Tools
Model APIs, open-weight model releases, developer documentation, and chat products.
-
Revenue Model
Usage-based API pricing and AI application access.
Key Information
-
Business Type
Private AI research lab
-
Headquarters
Hangzhou, Zhejiang, China
-
Founded Date
2023
-
Company CEO
Liang Wenfeng
-
Founders
Liang Wenfeng
-
Brands
DeepSeek, DeepSeek-V3, DeepSeek-R1
-
Categories
News
-
Employee Count
N/A
-
Funding
Backed by High-Flyer-related resources
-
Valuation
N/A
-
Ownership
Private
-
Stock Ticker
N/A
Latest DeepSeek News
DeepSeek Helps Shape Huawei’s Next AI Chips, Report Says
DeepSeek is reportedly working with Huawei on chips tailored to its AI models, alongside an expanding software ecosystem for Ascend processors.
US Orders DeepSeek Removed From Civilian Federal Agency Systems
DHS requires civilian federal agencies to remove DeepSeek and third-party tools built on its technology, while barring further procurement.
DeepSeek Updates V4.1 Flash and Opens Search APIs for AI Agents
DeepSeek's V4.1-Flash-1017 claims faster, shorter responses at the same model price while Search and Fetch APIs open its web tools to developers.
DeepSeek Nears $12 Billion Funding Round Backed by Tencent and CATL
DeepSeek is reportedly nearing a $12 billion funding round backed by Tencent and CATL, exceeding its initial target ahead of a planned IPO.
DeepSeek Launches Harness, a Desktop Rival to Claude Code
DeepSeek Harness brings a plugin-based AI agent to Mac and Windows, with coding tools, everyday workflows and extensions users can build through chat.
Xiaomi Releases MiMo V2.6 Pro and Flash, Topping the Open Weight Model Rankings
Xiaomi released MiMo-V2.6-Pro and Flash under an MIT license, with Pro becoming the top scoring open weight model on Artificial Analysis' Intelligence Index at a fraction of frontier lab training costs.
DeepSeek’s New V4.1 Flash is Already Challenging the Best
DeepSeek launched V4.1 Flash, an open-weight model on a new architecture that beats its own larger flagship and rivals Claude Opus 5 on some benchmarks at a fraction of the price.
Anthropic Says Moonshot Secretly Routed Kimi Requests to Claude
Anthropic alleges that Moonshot used thousands of Claude accounts to serve some Kimi responses, raising questions about model provenance and customer data handling.
US Agencies Accuse Six Chinese AI Firms of Systematic Model Theft
The FBI, NSA and CISA jointly accused six Chinese AI companies of systematically distilling US models including Claude and ChatGPT since 2024, calling it their core development strategy.