Latest open artifacts (#25): DeepSeek v4.1 Flash 1017, American-made models

China's Ministry of Commerce (MOFCOM) released a policy proposal that would require frontier AI models and some other technologies to obtain government pre-approval before being shared internationally, while the U.S. Department of Homeland Security issued an order barring DeepSeek as a company and its models as a technology platform from federal civilian procurement. DeepSeek separately published its internal inference stack, DeepInfer, on GitHub, which estimates cited by the newsletter suggest lets DeepSeek earn up to 90% gross margin on token sales. Xiaomi's MiMo-V2.6-Pro-RL, trained with heavier reinforcement learning on the MiMo V2.5 base, is now the top open model on Artificial Analysis's ranking.

The geopolitics of open-weight AI are accelerating again this fall, as we would expect with the risks of all AI systems open and closed rising rapidly. We’ve seen many news items in the last week showing both the U.S. and Chinese governments positioning themselves for some controls on open models. 1. MOFCOM, the Ministry of Commerce of the People’s Republic of China, released a policy proposal it’s not yet a law that would require frontier AI models https://x.com/zephyr z9/status/2111071903469478291 and some other technologies to get pre-approval from the government before they can be shared internationally. We found the reaction to this to be overblown – as the document states that open-source technologies will be set up for automatic approvals https://x.com/ZilanQian/status/2111073916194013274 – but many in geopolitical circles say this is a big deal because it proposes a regulatory framework that could be used in the future to regulate China’s AI industry. Others we trust on this issue say this proposal could be more about preventing issues like the sale of Chinese startup Manus https://www.npr.org/2026/04/27/g-s1-118892/china-blocks-meta-from-acquiring-ai-startup-manus to Western companies, rather than being about frontier AI models. Either way, the wheels of control are moving around AI, and open-weight models are not free from any influence. 2. The Department of Homeland Security DHS released an order on DeepSeek https://public-inspection.federalregister.gov/2026-21447.pdf , which restricts DeepSeek itself, as a company, and DeepSeek’s models, as a technology platform, from being a vendor in federal civilian procurement. It is not clear how this will relate to finetuned models or models trained on the outputs of DeepSeek e.g., Nvidia’s Nemotron models have long relied heavily on DeepSeek models in their training recipes – will they be banned? , but this is an expected action and reinforces the need for a domestic open model ecosystem. Other model makers in China, such as Zhipu AI, makers of GLM, are already on the Entity List under Commerce/BIS, which restricts their access to American technology. DeepSeek followed the above news up by releasing their internal inference stack publicly, DeepInfer https://github.com/deepseek-ai/DeepInfer , which is likely far more efficient than existing open-source tools like vLLM/SGLang for their models. Estimates https://x.com/zephyr z9/status/2112499092031185134?s=46 point to DeepSeek making up to 90% gross margin when selling tokens on this inference stack. It is well-known that the top AI labs have much more inference-efficient tools internally, and this is the first time we’ve seen a leading lab share them. Inference companies will be quick to adopt these tools, which will drive prices for DeepSeek’s models down across the ecosystem or margins up . While not an open model or dataset, we consider this release to be one of the bigger open-source AI artifacts in the last few months. As of writing this, Kimi K3.1 is rumored to be released very soon, with metadata associated with it starting to appear in public libraries https://x.com/mylifcc/status/2112397528272695299 . Exciting times You can view the models in this email, and all previous issues, on our Artifacts Hub. It includes commentary and other ecosystem metadata. Our Picks - MiMo-V2.6-Pro-RL https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL by XiaomiMiMo https://huggingface.co/XiaomiMiMo : an update for Mimo, which reuses the same bases as MiMo V2.5, but used a lot more resources on RL training. They also release an in-depth tech report https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL/blob/main/MiMo V2 6 technical report.pdf , as well as some of the used environments, which is rare at this scale. The final model from this recipe, V2.6-Pro, is the top open model on Artificial Analysis’s ranking today, and the scaled RL runs came with a fun live dashboard https://mimo.xiaomi.com/rl/ of training. We’ve been fans of Xiaomi for months now, and their efforts are bearing fruit. - DeepSeek-V4.1-Flash-1017 https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017 by deepseek-ai https://huggingface.co/deepseek-ai : Since the latest Artifacts, the whale has not released one, but two new model iterations with a new size 552B-A16B which also support image inputs. The model has gained a lot of well-deserved attention for its performance at a small price. It also carries the MIT license, continuing the trend established by DeepSeek themselves. This is notable in times where licenses seem to become more bespoke and exclusionary. Below is the Pareto of performance vs price on DeepSWE, a favorite agentic software engineering benchmark today, which DeepSeek dominates. - clef https://huggingface.co/Cloudflare/clef by Cloudflare https://huggingface.co/Cloudflare : Decision models are all the rage after being introduced by the startup Typesafe https://typesafe.ai/blog/introducing-system-one-models-and-jev . In short, they are general classification models that output few tokens with their probabilities. Its usefulness has led to quick adoption and also a lot of open model finetunes such as Clef. - NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 by nvidia https://huggingface.co/nvidia : The mid-sized model of the Nemotron family got its long awaited update. It is strong for its size and can be run at home given you own a beefy laptop or RTX Spark . As usual, NVIDIA releases the vast majority of the data and an in-depth tech report https://research.nvidia.com/labs/nemotron/files/NVIDIA-Nemotron-3.5-Super-and-Lightning-Technical-Report.pdf . - Beam-1.0-501B-A23B https://huggingface.co/reflection/Beam-1.0-501B-A23B by reflection https://huggingface.co/reflection : The long awaited open model by Reflection, making it the strongest open American model released under a permissive license Apache 2.0 . While not the smartest open model available, Beam is very impressively token efficient second to only GPT-6-Astra below . This is a sign of a well-developed post-training stack. Models General Purpose - Ternary-Bonsai-2-27B-gguf https://huggingface.co/prism-ml/Ternary-Bonsai-2-27B-gguf by prism-ml https://huggingface.co/prism-ml : A ternary version of Qwen3.8-27B for on-device deployments. - Ornith-1.5-397B https://huggingface.co/ornith-ai/Ornith-1.5-397B by ornith-ai https://huggingface.co/ornith-ai : A series of Qwen3.5 and Gemma 4 finetunes that gained attention in the local AI community for its strong performance over the respective strong baselines. - Ling-3.0-flash-VL-fp4 https://huggingface.co/inclusionAI/Ling-3.0-flash-VL-fp4 by inclusionAI https://huggingface.co/inclusionAI : A 124B-A5.5B model by Ant Ling within Ant Group , which continues to churn out models consistently. - Muse-Spark-Open-816B-A42B https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B by meta-models https://huggingface.co/meta-models : META has released their open version of Muse, which we have written about in more detail https://www.interconnects.ai/p/meta-shakes-up-the-frontiers-open