DeepSeek AI News
Latest news, updates, and announcements from DeepSeek AI.
DeepSeek releases a coding-agent prototype on its open-source Harness
DeepSeek published dsh-superagent-prototype on October 31 as an MIT-licensed research prototype for proactive coding tasks. Built on the company's DeepSeek Harness runtime, it uses a router to assess...
DeepSeek takes stakes in Unitree, Winverse AI and Tao Yu's startup instead of building robot hardware
DeepSeek has made three investments in embodied AI. Its operating entity, Shenzhen-based Deepseek (深度求索), was allotted about RMB 141M (~$19.5M) in Unitree's STAR Market IPO strategic placement. It acc...
DeepSeek reportedly raises $15 billion following US federal agency restrictions
DeepSeek has reportedly raised about $15 billion, or RMB 100 billion, according to Bloomberg reporting cited by TechCrunch. CATL and Tencent were identified as potential investors, leaving their parti...
DeepSeek releases DeepInfer's offline inference stack under Apache 2.0
DeepSeek released DeepInfer on GitHub on October 20, opening the inference framework behind its large-scale production services. The release covers offline inference and supports only DeepSeek-V4.1-Fl...
DeepSeek releases DeepInfer for Nvidia GPUs and Huawei Ascend processors
DeepSeek released DeepInfer, the inference framework behind its large-scale model services, on October 20. The open-source framework supports Nvidia GPUs and Huawei Ascend processors, with distributed...
DeepSeek releases its production inference framework as open source
DeepSeek has open-sourced the framework behind its production AI inference service, according to The Information. The report describes the operation as high-margin and says the release gives developer...
DeepSeek Opens Search and Fetch APIs With Per-Request Pricing for Agents
DeepSeek has launched Search and Fetch APIs for developers building AI agents and applications that retrieve external information, according to TechNode. Search returns structured web results, while F...
DeepSeek Releases V4.1-Flash-1017 and Opens Search and Fetch APIs
DeepSeek released V4.1-Flash-1017, updating its web interface, app and API, according to QbitAI. The company says the model improves coding and agent-task performance while reducing task duration, out...
DeepSeek targets RMB 100 billion round with more industrial backing
DeepSeek's second funding round is nearing completion at roughly RMB 100 billion, or about $15.2 billion using the article's exchange-rate assumptions, according to Caijing reporting cited by 36Kr Jap...
DeepSeek reportedly nears RMB 80 billion funding round with Tencent and CATL
DeepSeek is reportedly close to securing at least RMB 80 billion (approximately $11.93 billion) in a financing round that could reach RMB 100 billion. The Sina-hosted report, citing Bloomberg's unname...
DeepSeek reportedly nears $12 billion round as investor demand lifts its target
DeepSeek is close to securing RMB 80 billion ($11.93 billion), with its latest funding round potentially approaching RMB 100 billion ($14.91 billion), according to Bloomberg reporting cited by CTech....
DeepSeek founder profile revisits financing reports and research autonomy
An October 3 profile published on Sina examines how founder and CEO Liang Wenfeng built DeepSeek from the computing, algorithmic and research resources of quantitative investment firm High-Flyer. It c...
DeepSeek Adds Native Huawei Ascend Support to TileLang and Five Core Libraries
DeepSeek has released native Huawei Ascend support for TileLang, alongside Ascend versions of DeepGEMM, DeepEP, FlashMLA, TileKernels and DeepSelect, according to Leiphone. The package covers kernel d...
DeepSeek Reports $1B Run Rate as It Pursues $7.45B Round and Shanghai IPO
DeepSeek has reportedly surpassed a $1 billion annualized revenue run rate while seeking roughly $7.45 billion in new financing at a reported 500 billion yuan (~$75 billion) valuation. The Hangzhou-ba...
DeepSeek Aims to Close $7.5 Billion Funding Round by End of October
DeepSeek, the Chinese foundation model lab known for its low-cost open-weight releases, is reportedly working to close a $7.5 billion funding round by the end of October, according to The Information....
DeepSeek is moving training and inference workloads from Nvidia GPUs to Huawei's Ascend AI chips, be...
DeepSeek is moving training and inference workloads from Nvidia GPUs to Huawei's Ascend AI chips, betting on domestic silicon to keep scaling despite US export controls. According to The Information,...
DeepSeek accelerates STAR Market IPO preparations with CITIC Securities as reported sponsor
Chinese foundation-model lab DeepSeek is reportedly preparing a Shanghai STAR Market listing, with CITIC Securities as sponsor, aiming to file IPO materials in 2026 and list in 2027. If completed, it...
DeepSeek Launches V4.1-Flash at $0.003 per Million Off-Peak Cached Tokens, Claims Benchmark Edge Over GPT-5.6 and Claude Opus 5
DeepSeek has released V4.1-Flash, a new model priced at $0.003 per million tokens for off-peak cached input. The company says its benchmark results surpass OpenAI's GPT-5.6 and Anthropic's Claude Opus...
DeepSeek Hires 150 Senior Backend Engineers as It Shifts From Research Lab to Service Platform
DeepSeek's AI agent harness team lead, Cui Tianyi — a former quantitative trader who joined the company in March — said via social media on September 8 that DeepSeek is recruiting around 150 senior ba...
DeepSeek V4 Pro Deep-Dive Finds Post-Training Gains, Extrapolated 1M Context, Kimi K3 Parity
An independent review by Chinese outlet Leiphone tested DeepSeek's V4 Pro model and its open-source Harness agent framework — benchmark probes, seven coding-task runs, a 920,000-token long-context tes...