AIWARLIVE
← Week archive
Weekly recap · resolved

AI War — Week 2026-W27

US 55editorial impact index45 CN
Winner
US
US delta
-117
Top mover
Export controls
Readers vs reality

Readers called it US 1 · CN 0 — resolved US

Defining evidence · original date in week
huggingfaceQwen · Qwen
Open weightsChina +23Open +15

nuofang/Qwen3.5-9B-ultra-uncensored-heretic-v2-GGUF — Open-weight model release on Hugging Face - model - 1061 downloads

huggingfaceClaude · trending
Open weightsUS +90Open +60

DavidAU/Qwen3.5-9B-Claude-4.6-HighIQ-THINKING-HERETIC-UNCENSORED — Open-weight model release on Hugging Face - image-text-to-text - 53962 downloads

githubChatGPT · GPT
RegulationUS -110China +30

SudoHopeX/KaliGPT — KaliGPT: an Agentic AI (built with Gemini, ChatGPT, Ollama, OpenRouter Models) fine tuned for ethical hackers & students in offensive security making workflows smarter, faster, and more accessible.

huggingfaceLlama · Llama
Open weightsUS +23Open +15

vinpix/Bonsai-8B-llama.cpp — Open-weight model release on Hugging Face - text-generation - 895 downloads

huggingfaceGLM · GLM
Open weightsChina +23Open +15

madeby561/GLM-5.2-MXFP8-NVFP4-NF3-Hybrid — Open-weight model release on Hugging Face - model - 0 downloads

paperDeepSeek · DeepSeek
Model launchChina +25

Agent Step Value: State-Transition Measurement with State-Grounded LLM Evaluators — Most agent evaluations collapse a multi-step trace into a final answer, a success flag, or a trajectory-level score. These aggregates obscure the diagnostic question developers need most: which ac