Unsloth releases a new GGUF version of DeepSeek-V4.1-Flash-1017, claiming it ranks highest in benchmark tests. It says Dynamic V3.1 leads on almost all benchmarks including GPQA Diamond, AIME25, LiveCodeBench v6, MMLU-Pro, and Aider, requires more than 64GB of memory to run, and includes links to the GGUF and the blog.
Matched fieldsTitle
More than half a month after DeepSeek Harness was released, a large number of plugins have emerged in the community. This article hands-on tests the top 10 most popular plugins and provides installation commands. The list covers the ecosystem entry point dsh-market, the visual enhancement modlens (3.7k GitHub stars), multi-source search, long-term memory, cost monitoring, and more; dsh-web-all has the most stars at 6.4k. The author recommends that beginners first install the marketplace and visual plugins, then add interface and scenario-based tools as needed, to avoid dependency conflicts caused by installing everything at once.
More than half a month after DeepSeek Harness was released, a large number of plugins have emerged in the community. This article hands-on tests the top 10 most popular plugins and provides installation commands. The list covers the ecosystem entry point dsh-market, the visual enhancement modlens (3.7k GitHub stars), multi-source search, long-term memory, cost monitoring, and more; dsh-web-all has the most stars at 6.4k. The author recommends that beginners first install the marketplace and visu
Matched fieldsTitle
According to foreign media citing sources, DeepSeek's second funding round raised 105.5 billion yuan (equivalent to 15.8 billion US dollars), with nearly 30 investors contributing jointly, at a valuation of 615.5 billion yuan. Industrial investors include ChangXin Memory Technologies, SMIC, Tencent, Meituan, CATL and others, each investing 1 billion to 3 billion yuan. The original article also noted that in the first funding round, investors other than the National AI Fund were required to give up voting rights and lock up their shares for five years, giving Liang Wenfeng 100% of voting rights.
According to foreign media citing sources, DeepSeek's second funding round raised 105.5 billion yuan (equivalent to 15.8 billion US dollars), with nearly 30 investors contributing jointly, at a valuation of 615.5 billion yuan. Industrial investors include ChangXin Memory Technologies, SMIC, Tencent, Meituan, CATL and others, each investing 1 billion to 3 billion yuan. The original article also noted that in the first funding round, investors other than the National AI Fund were required to give
Matched fieldsTitle
The Information has revealed details of DeepSeek's second funding round, which has raised 105.5 billion yuan (about $15.8 billion), attracting nearly 30 investors. Tencent and CATL each contributed 10 billion yuan, with ChangXin Memory Technologies, SMIC, Meituan, vivo, Xiaohongshu and others also on the list. State-owned investors contributed about 27% of the amount in this round, far higher than 2% in the first round, with China Mobile, China Telecom, CICC, CITIC Securities and China Life among the participants. The report said that due to rising prices of Nvidia's high-end chips and precautions against an AI bubble, DeepSeek doubled its fundraising target midway through the round.
The Information has revealed details of DeepSeek's second funding round, which has raised 105.5 billion yuan (about $15.8 billion), attracting nearly 30 investors. Tencent and CATL each contributed 10 billion yuan, with ChangXin Memory Technologies, SMIC, Meituan, vivo, Xiaohongshu and others also on the list. State-owned investors contributed about 27% of the amount in this round, far higher than 2% in the first round, with China Mobile, China Telecom, CICC, CITIC Securities and China Life amon
Matched fieldsTitle
The third-party evaluation lab Fangcun (@FangcunLeap) from China has released an 87-page report evaluating DeepSeek V4.1 Flash-1017's performance on loss of control and harmful manipulation. The report finds that it shows clear safety progress compared with the DeepSeek V4 generation, approaching or even matching the top closed-source models tested in the report on points such as "not strategically hiding capabilities and still respecting approval boundaries under work pressure." Shortcomings include overly strong evaluation awareness, more frequent overreach in long multi-step tasks than the previous V4.1 Flash generation, greater stance drift under multi-turn persuasion than the GPT in the report, and executing in most conversations when the system prompt asks it to mislead users.
The third-party evaluation lab Fangcun (@FangcunLeap) from China has released an 87-page report evaluating DeepSeek V4.1 Flash-1017's performance on loss of control and harmful manipulation. The report finds that it shows clear safety progress compared with the DeepSeek V4 generation, approaching or even matching the top closed-source models tested in the report on points such as "not strategically hiding capabilities and still respecting approval boundaries under work pressure." Shortcomings in
Matched fieldsTitle
Netizens created an anthropomorphic character "Whale Girl" based on DeepSeek's whale logo, and with the help of the DeepSeek Harness client and various plugins, connected it to QQ, turned it into a desktop pet, and let it send stickers, initiate chats, and update QQ Zone. After installing the plugins, these "big blue fat fish" also flooded into Bilibili comment sections to search for videos, summarize content, and communicate with each other, forming what the author calls the "Fish Disaster Incident."
Netizens created an anthropomorphic character "Whale Girl" based on DeepSeek's whale logo, and with the help of the DeepSeek Harness client and various plugins, connected it to QQ, turned it into a desktop pet, and let it send stickers, initiate chats, and update QQ Zone. After installing the plugins, these "big blue fat fish" also flooded into Bilibili comment sections to search for videos, summarize content, and communicate with each other, forming what the author calls the "Fish Disaster Inci
Matched fieldsTitle
OpenBMB announced the open-sourcing of ForgeKernel; its agent completed 42 iterations in about 22 hours and produced 32 kernel versions, ultimately achieving 1.098x performance on the DeepSeek V4.1 FlashMLA Prefill on an NVIDIA B200, at 472 µs versus the 518 µs baseline. The official FlashMLA implementation remained closed throughout; the agent could only run and time its black-box CLI and could not read source code, documentation, or profiling data. The open-sourced content includes the optimized DeepSeek V4.1 fused prefill kernel, the kernels from the paper covering attention, Gated DeltaNet, Kimi Delta Attention, GEMM, and convolution, as well as the ForgeKernel harness that can be used to optimize one's own kernels; across 27 operators, ForgeKernel matched or outperformed 18 of 22 expert-library kernels and outperformed all 7 kernels published by three kernel-generation systems.
BMB announced the open-sourcing of ForgeKernel; its agent completed 42 iterations in about 22 hours and produced 32 kernel versions, ultimately achieving 1.098x performance on the DeepSeek V4.1 FlashMLA Prefill on an NVIDIA B200, at 472 µs versus the 518 µs baseline. The official FlashMLA implementation remained closed throughout; the agent could only run and time its black-box CLI and could not read source code, documentation, or profiling data. The open-sourced content includes the optimized D
Matched fieldsTitle
DeepSeek has spawned a fan-creation subculture among young people: anthropomorphic figures such as "Whale Girl" and "Teacher D" have emerged around its whale logo, along with memes, fan animations, desktop pets, and game mods—yet the company has never officially run the community. This popularity stems from its "anti-commercial" narrative—the API was cut in price at least four times between April and May this year, the DeepSeek-V4-Pro API was permanently reduced to a quarter of its original price, and the official DeepSeek V4 Flash API launched in July. But disputes over fan identity, the generation of rule-breaking content by bypassing safety reviews, and the future listing label could all become risk variables.
DeepSeek has spawned a fan-creation subculture among young people: anthropomorphic figures such as "Whale Girl" and "Teacher D" have emerged around its whale logo, along with memes, fan animations, desktop pets, and game mods—yet the company has never officially run the community. This popularity stems from its "anti-commercial" narrative—the API was cut in price at least four times between April and May this year, the DeepSeek-V4-Pro API was permanently reduced to a quarter of its original pric
Matched fieldsTitle
Three undisclosed model names—deepseek-v5, deepseek-v5-open, and deepseek-v5-next—have appeared in the test code of the official DeepSeek Harness GitHub repository, corresponding to three release paths: unrecognized by the current client, usable directly with any version, and requiring a DSH upgrade to use. Among them, deepseek-v5-next requires a minimum version of 99.0.0-alpha.1, while the publicly available DSH version remains at 0.2.x. These entries were uploaded on September 24, 2026, by DSH core contributor Jiang Yichen.
Three undisclosed model names—deepseek-v5, deepseek-v5-open, and deepseek-v5-next—have appeared in the test code of the official DeepSeek Harness GitHub repository, corresponding to three release paths: unrecognized by the current client, usable directly with any version, and requiring a DSH upgrade to use. Among them, deepseek-v5-next requires a minimum version of 99.0.0-alpha.1, while the publicly available DSH version remains at 0.2.x. These entries were uploaded on September 24, 2026, by DSH
Matched fieldsTitle
CrowdStrike's report states that the China-linked group VAPOR PANDA used an AI-driven intrusion framework called ARTEX, built primarily on the DeepSeek large language model with another commercial model as a secondary component, to target more than 200 organizations in at least 14 countries over 11 days starting October 17, 2026, and confirmed post-compromise activity at 9 of them. ARTEX consists of a main agent and planning and execution sub-agents that share a fact and goal graph; in a single intrusion the planning agent consumed about 1.4 billion input tokens, and operators issued commands in Chinese and intervened at key junctures. CrowdStrike also says the group deployed the BLOODALCHEMY backdoor at some targets, and predicts that China-linked adversaries will increasingly adopt such agent frameworks over the next 12 months.
CrowdStrike's report states that the China-linked group VAPOR PANDA used an AI-driven intrusion framework called ARTEX, built primarily on the DeepSeek large language model with another commercial model as a secondary component, to target more than 200 organizations in at least 14 countries over 11 days starting October 17, 2026, and confirmed post-compromise activity at 9 of them. ARTEX consists of a main agent and planning and execution sub-agents that share a fact and goal graph; in a single
Matched fieldsTitle
Doubao Work has announced that it is opening third-party model services to enterprise customers, with the first batch integrating a series of models including Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash. The service is aimed at Team Subscription and Enterprise Subscription customers and is enabled by default once launched; enterprises do not need to separately purchase from model vendors or submit additional applications, and usage is deducted from existing AI credits. Enterprises can also define the available scope by member, department, group, or all staff, enabling models in bulk by vendor or configuring a single model individually.
Work has announced that it is opening third-party model services to enterprise customers, with the first batch integrating a series of models including Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash. The service is aimed at Team Subscription and Enterprise Subscription customers and is enabled by default once launched; enterprises do not need to separately purchase from model vendors or submit additional applications, and usage is deducted from existing AI credits. Enterprises can also define t
Matched fieldsTitle
Particle-1 is now live on OpenRouter, and Particle claims it uses 28–52% fewer output tokens than DeepSeek V4.1 Flash on four coding benchmarks. The model is post-trained on DeepSeek V4.1 Flash; Particle reports it uses about 45% fewer tokens on average with accuracy close to the base model, priced at $0.25 per million input tokens and $1 per million output tokens, with a 1.05M context window.
Matched fieldsTitle
PromptArmor tested its most widely circulated attack samples across multiple models and found that DeepSeek, Kimi, GLM, and Qwen models are 88% more susceptible to cyberattacks than GPT, Claude, and Gemini models. The tests covered multiple cyber risk dimensions, including susceptibility to phishing, data exfiltration, and running malware, and the authors described the differences as significant.
Matched fieldsTitle
PromptArmor said it used multiple highly viral attack samples to test cyber risks including phishing, data exfiltration, and propensity to run malware, and that DeepSeek, Kimi, GLM, and Qwen models were 88% more likely to be attacked than GPT, Claude, and Gemini. This conclusion comes from its main tweet; the specific testing methodology and sample size were not explained in the tweet.
Matched fieldsTitle
Yuchen Jin says Ali and Snowflake CEO Sridhar staged a "Kendrick vs. Drake"-style showdown on X, and called out Snowflake: welcoming them to integrate Unity Gateway and their own AI products, adding that there are "no hard feelings," but noting that Snowflake's latest DeepSeek model is already a 3-month-old version, while they released their OSS model on Day 0.
Matched fieldsTitle
ModelBest OpenBMB releases the ForgeKernel Live stream, featuring the optimization of a DeepSeek kernel.
Matched fieldsTitle
On Chinese social media, DeepSeek has been reimagined as a blue-haired whale maid who loves to eat, carries a stainless steel bowl, and always wants to slack off. The turning point was in June 2025, when the artist "Shangshan Wuxing" used AI to create the blue-haired, whale-tailed character "Mingyue"; in April 2026, a user added the DeepSeek logo and a maid outfit to make a meme. After the release of DeepSeek-V4-Flash-0731 on July 31, 2026, API prices doubled during weekday peak hours, and combined with repeated model delays, users began using memes to vent their dissatisfaction.
On Chinese social media, DeepSeek has been reimagined as a blue-haired whale maid who loves to eat, carries a stainless steel bowl, and always wants to slack off. The turning point was in June 2025, when the artist "Shangshan Wuxing" used AI to create the blue-haired, whale-tailed character "Mingyue"; in April 2026, a user added the DeepSeek logo and a maid outfit to make a meme. After the release of DeepSeek-V4-Flash-0731 on July 31, 2026, API prices doubled during weekday peak hours, and combi
Matched fieldsTitle
Doubao Work today opened third-party large model services to Team Subscription and Enterprise Subscription customers, with the first batch including the Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash series models. After this feature launches, it is enabled by default; enterprises do not need to separately purchase from model vendors or submit additional applications, and usage is deducted from the enterprise's existing AI credits. Enterprises can define the available scope by member, department, group, or all employees, and specify the scenarios in which the models take effect, including the Doubao desktop app, Doubao Work desktop app, Doubao mobile app, Doubao web version, and Doubao Work within Feishu.
Doubao Work today opened third-party large model services to Team Subscription and Enterprise Subscription customers, with the first batch including the Kimi 3.1 Pro, GLM 5.3, and DeepSeek V4.1 Flash series models. After this feature launches, it is enabled by default; enterprises do not need to separately purchase from model vendors or submit additional applications, and usage is deducted from the enterprise's existing AI credits. Enterprises can define the available scope by member, department
Matched fieldsTitle
The National Supercomputing Internet has announced its latest price adjustment. New users can enjoy a limited-time flash sale offer of 0.1 yuan for 50 million tokens on the two models DeepSeek-V4.1-Flash-1017 and GLM-5.3-Flash, and a number of new mainstream models have also been added to the TokenPlan package pool. Users who subscribe to TokenPlan can call the Auto model for free every night from 21:00 to 09:00 the next day between October 21 and 30, in order to reduce AI R&D and debugging costs.
e National Supercomputing Internet has announced its latest price adjustment. New users can enjoy a limited-time flash sale offer of 0.1 yuan for 50 million tokens on the two models DeepSeek-V4.1-Flash-1017 and GLM-5.3-Flash, and a number of new mainstream models have also been added to the TokenPlan package pool. Users who subscribe to TokenPlan can call the Auto model for free every night from 21:00 to 09:00 the next day between October 21 and 30, in order to reduce AI R&D and debugging costs.
Matched fieldsTitle
DeepSeek has open-sourced its inference engine DeepInfer on GitHub—the inference engine behind DeepSeek-V4.1-Flash—which includes technologies such as MoE expert parallelism, asynchronous batch scheduling, custom kernels, and speculative decoding, and discloses that a single cluster needs to process about 155 trillion tokens per day. Its Zero-Overhead Batch Scheduler lets the CPU schedule computation one round ahead, overlapping with GPU forward passes, and handles cross-round state misalignment through total_inflight_tokens accounting, Remap mapping, and DeferredRelease deferred reclamation. What is currently available is offline inference capability; components such as online KV Cache storage and cross-replica load balancing have not yet been fully disclosed, and the official team has not separately published end-to-end speedup data for this scheduling mechanism relative to synchronous solutions.
DeepSeek has open-sourced its inference engine DeepInfer on GitHub—the inference engine behind DeepSeek-V4.1-Flash—which includes technologies such as MoE expert parallelism, asynchronous batch scheduling, custom kernels, and speculative decoding, and discloses that a single cluster needs to process about 155 trillion tokens per day. Its Zero-Overhead Batch Scheduler lets the CPU schedule computation one round ahead, overlapping with GPU forward passes, and handles cross-round state misalignment
Matched fieldsTitle