English text is being prepared.
LatestUnsloth releases GGUF version of DeepSeek-V4.1-Flash-1017
Why read thisUnsloth has released a GGUF quantized version of DeepSeek-V4.1-Flash-1017, along with a 64GB memory requirement, making it easier for local deployers to evaluate.
According to The Information, DeepSeek's second funding round has raised 105.5 billion yuan (approximately 15.8 billion US dollars), attracting nearly 30 investors, with state-owned capital contributing about 27%, a sharp increase from 2% in the first round. Tencent invested 10 billion yuan in each of the two rounds; CATL and its chairman Zeng Yuqun invested 5 billion yuan and 10 billion yuan in the two rounds respectively; and industrial parties such as ChangXin Memory Technologies, SMIC, vivo, Meituan, Geely, Midea, Luxshare Precision, Trip.com, and Xiaohongshu are also among the investors. The report, citing people familiar with the matter, said that rising prices for Nvidia's high-end chips and concerns about an AI bubble were the reasons DeepSeek doubled the size of its fundraising midway.
Why read thisThe funding details disclose the composition of investors and changes in the share of state-owned capital, which can be used to observe DeepSeek's capital structure and the extent of industrial-party participation.
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.
Why read thisThe CrowdStrike report discloses the division of labor and scale of an AI agent framework in a real intrusion, offering insight into how automated attacks operate.
Why read thisPromptArmor ran a cross-model comparison using its own widely circulated attack samples, revealing differences between Chinese and foreign models in phishing, data exfiltration, and running malware.
Anthropic publicly filed its S-1 on Monday, planning to list on Nasdaq under the ticker ANT, with revenue growing from $41 million in 2023 to $16.3 billion in the first half of 2026, including $11.6 billion in Q2 alone, up 1368% year over year. Q2 operating profit was $974 million, first-half operating cash flow was positive at $3.9 billion, while the $80.7 billion net loss mainly came from $81.1 billion in non-cash financial instrument expenses. About 48% of first-half revenue (roughly $7.8 billion) flowed through Bedrock, Vertex, and Azure Marketplace; the S-1 also disclosed that pretraining of a frontier model was recently delayed, and listed NVIDIA compute commitments of up to $120.5 billion.
Why read thisThe original article summarizes the key financial figures from Anthropic's S-1 and reactions from various parties, so readers can understand its revenue structure, compute commitments, and points of controversy.
Why read thisThe author uses price comparisons and benchmark performance to show the impact of open-source models on the closed-source pricing system.
Why read thisKimi K3.1 Pro's Code Arena ranking and cost data, directly comparable with Claude Opus-5 and the previous-generation K3.
Moonshot AI has released Kimi K3.1 Pro, an upgraded version of Kimi K3, featuring a MoE architecture with 2.8 trillion total parameters and 104 billion activated parameters, activating 16 of 896 experts at a time. According to Artificial Analysis benchmarks, the model ranks fourth on the Intelligence Index with a score of 55, higher than GPT-6 Astra's 53, while the previous generation Kimi K3 scored 44. It ranks first on Terminal Bench 4.0 with 67%, compared with 13% for the previous generation and 64% for Claude Sonnet 5.5; pricing is $2 per million input tokens and $10 per million output tokens, but its output speed of 73 tokens/s is below the average of 81 tokens/s.
Why read thisThe leaderboard provides sub-item scores and per-task cost, allowing comparison of the gap between open-weight models and restricted-access models on cybersecurity tasks.
Moonshot AI released the reasoning model Kimi K3.1 Pro, which scores 55 on the Artificial Analysis Intelligence Index, ranking 4th, an 11-point improvement over Kimi K3's 44 points, trailing only Claude Opus 5.5 (58), Claude Sonnet 5.5 (56), and Muse Wonder Full (56). Its agent capabilities improved notably: it scores 1,846 Elo on GDPval-AA, ranking 3rd; 67% on Terminal-Bench 4.0 (Kimi K3 was 13%); and AutomationBench-AA rose from 58% to 73%. The model has a 1M-token context window, 2.8T total parameters and 104B activated parameters, supports text and image input, and is priced at $2/10/0.20 per 1M input/output/cache-hit tokens, all three 33% lower than Kimi K3; the weights have not yet been released, and once released it will become the highest-scoring open-weight model, narrowing the gap between leading open-source and closed-source models from 12 points to 3 points.
Why read thisThe evaluation organization provides specific scores for Kimi K3.1 Pro on the Intelligence Index and agent benchmarks, and compares its gap with leading closed-source models.
Why read thisARC Prize provides per-version scores and per-task costs for the DeepSeek Flash series on ARC-AGI, enabling horizontal comparison of model iteration magnitude.
Recursive, founded by Tian Yuandong, has open-sourced its first model Flint, focused on local Agent coding, with 40 billion total parameters, 2.5 billion activated per token, and support for a 512k token context window. The company says that paired with its self-developed inference engine rec, Flint can reach at least 125 tokens/s on a MacBook Pro and over 150 tokens/s on RTX Spark, roughly 1.4–2.6 times higher throughput than llama.cpp. Flint scores 74.2% on SWE-bench Verified, 89.7% on SWE-bench Pro v2, and 54.9% on Terminal-Bench 2.1; rec currently supports M4 Pro, M5 Pro, and M5 Max, requiring at least 48GB of unified memory and about 30GB of free disk space.
Why read thisTian Yuandong's startup Recursive open-sources its first model, letting readers understand the trade-offs of local Agent coding models in speed and coding benchmarks.
According to Bloomberg, DeepSeek's second funding round is locked at about 100 billion yuan, with a pre-money valuation of about 500 billion yuan; the fundraising target was raised twice from the initial 50 billion yuan and then doubled. Existing shareholders Tencent and CATL each added about 10 billion yuan; new entrants include industrial capital such as Shixi Capital, SMIC Capital, CPE Yuanfeng, Legend Capital and Geely. Perfect World, Century Huatong and Litong Electronics participate indirectly through two special funds, Lisi Xingshen and Shixiang Shangshi; DeepSeek is also preparing for a STAR Market IPO and has already invested in embodied intelligence companies such as Unitree.
Why read thisThe original text sorts out the investor structure and external fund channels of the hundred-billion financing, and gives three investment clues for the embodied intelligence layout.
Tomasz Tunguz converts the inference data disclosed by DeepSeek and SpaceX's third-quarter financial report into the same basis, yielding four price points for revenue per GPU card-hour, from DeepSeek's theoretical maximum of $38 to SpaceX's actual $2.44. The B200 spot price is $8.77, the SpaceX contract price is $5.27, and the payback period for its new compute deployment is less than one year. The author also points out reverse pressure: GPU rent rose from $4.40 to $8.08 within half a year, while the cost of the equivalent benchmark threshold fell 377 times in 18 months, and JPMorgan says more than 60% of the data center capacity planned for 2027 has not yet started construction.
Why read thisThe author converts DeepSeek's and SpaceX's data into revenue per card-hour on the same basis, so readers can understand the price space and cost pressure of the inference business.
Why read thisThe author estimates gross margin using publicly available per-GPU-hour revenue and GPU rental costs, providing a reference point for assessing the economics of inference services.
DeepSeek has open-sourced DeepInfer, its self-developed inference framework that supports DeepSeek-V4.1-Flash, and published production data: a single cluster processes 155 trillion tokens per day, 96% input cache hit rate, 400ms first-token latency, and 230 tokens/s generation speed. The framework is customized based on vLLM and supports distributed inference, speculative decoding, SWA Bounded Replay, and structured output, with target platforms NVIDIA SM100 and Huawei Ascend 950. DeepSeek also states that components such as KV cache management and load balancing are not yet fully open-sourced, while the vLLM team says that under a matching GB200 configuration, decoding throughput is within 6% of the results published by DeepInfer.
Why read thisThe original article puts DeepInfer's production data, cache tiering, and vLLM reproduction comparison together, making it easier to judge the transferable boundaries of this inference stack.
Why read thisThird-party benchmarks compare this model with Flash models in the same tier on scores, token consumption, and per-task cost, which can be used to assess its price-performance position.
Why read thisAgent Arena's empirical leaderboard shows that DeepSeek-V4.1-Flash has entered the top ranks at a lower cost, allowing readers to compare the cost-effectiveness of open-source and closed-source models.
Why read thisUnsloth provided the accuracy retention ratio and size change after 1-bit quantization, so readers can judge the local deployment threshold accordingly.