https://blog.lynxflow.co/en/tags/glm/Recent content in GLM on Lynx Tech BlogHugo -- gohugo.ioenFri, 11 Sep 2026 00:00:00 +0800https://blog.lynxflow.co/en/posts/best-llm-for-novel-writing-2026-09/Fri, 11 Sep 2026 00:00:00 +0800https://blog.lynxflow.co/en/posts/best-llm-for-novel-writing-2026-09/Picking an AI model for novel writing, the internet is full of claims — &ldquo;Claude has the best prose,&rdquo; &ldquo;DeepSeek has the densest foreshadowing,&rdquo; &ldquo;Kimi is in a league of its own for ultra-long-context continuation.&rdquo; Which of these are backed by actual testing, and which are marketing? This article pulls together all publicly available raw benchmark data as of September 2026, evaluates models across the five dimensions that actually matter for novel writing, and ghttps://blog.lynxflow.co/en/posts/pro20x-vs-china-llm/Mon, 17 Aug 2026 00:00:00 +0800https://blog.lynxflow.co/en/posts/pro20x-vs-china-llm/<img src="https://blog.lynxflow.co/images/pro20x-vs-china-llm.png?v=090818" alt="Featured image of post ChatGPT pro 20x Subscription vs China's Top LLMs: Per-Million-Token Cost" />A counterintuitive ledger Suppose you pay 1000 RMB to activate a ChatGPT &ldquo;pro 20x&rdquo; subscription, equivalent to roughly $1700 of API value per week, and stack OpenAI&rsquo;s ~10 monthly &ldquo;Goodwill Resets&rdquo; (announced by tibo — each one refreshes a brand-new fully-loaded 7-day cycle). Under that assumption, your effective cost per million tokens lands at 0.036 ~ 0.91 RMB — while China&rsquo;s cheapest DeepSeek-V4-Flash costs 1.06 RMB, ~1.2x higher; against the priciest Kimi K