{"slug": "kimi-k3-is-not-cheap", "title": "Kimi K3 is not cheap", "summary": "Moonshot AI's new open-weights LLM Kimi K3 is not cheap, costing slightly less per task than OpenAI's top model but double the cost of GLM-5.2 and about 20 times more than DeepSeek V4, according to Artificial Analysis's Intelligence Index benchmark. The model is a big step forward for Chinese LLMs and open-weights development, but claims that it is as good as US models at a fraction of the cost are incorrect.", "body_md": "# Kimi K3 Is Not Cheap\n\nThe new Chinese LLM is not what (some) people think.\n\nKimi K3, the new open-weights LLM from Chinese lab *Moonshot AI*, dropped last week. It’s a powerful model that, perhaps inevitably, sparked debate — including accusations of ushering in decelerationist ‘[AI communism](https://xcancel.com/deanwball/status/2078133895766114412?s=20)’ (whatever that means).\n\nAmidst the noise, a number of AI commentators (including the Hard Fork guys and Scott Galloway) have asserted that K3 is noteworthy because it’s nearly as good as US models, at a fraction of the cost.\n\nUnfortunately, they are wrong. Just because a model is Chinese, it is not magically cheap. And in fact, K3 isn’t cheap. Across Artificial Analysis’s ‘[Intelligence Index](https://artificialanalysis.ai/)’ benchmark, it cost slightly less per task than OpenAI’s top model. Compared to other Chinese models, it’s positively expensive - double the cost of GLM-5.2, and ~20x more than DeepSeek V4.\n\nThere’s nuance, of course. Kimi costs less for [coding](https://artificialanalysis.ai/agents/coding-agents), but more for [office work](https://artificialanalysis.ai/evaluations/aa-briefcase). It’s verbose, so it takes longer to solve problems. It’s open-weights, so you could (with a few GPUs) finetune it for your own purposes. But we’re not talking nuance here, this is just about one claim that I’ve seen/heard enough to find slightly annoying.\n\nIn summary: Is K3 a big step forwards for Chinese LLMs (and the open-weights people)? *Yes*. Is it concerning for the US market that a Chinese lab have fast-followed OpenAI and Anthropic with much less capex, possibly aided by distillation? *Also yes*. **Is it cheap?** *No. No it isn’t. Please stop saying that. And please stop saying it’s another ‘DeepSeek moment’ too.*\n\nPosted on July 26, 2026", "url": "https://wpnews.pro/news/kimi-k3-is-not-cheap", "canonical_source": "https://www.alexinch.com/blog/kimi-k3", "published_at": "2026-07-26 19:37:09+00:00", "updated_at": "2026-07-26 19:52:09.776801+00:00", "lang": "en", "topics": ["large-language-models", "artificial-intelligence", "ai-startups"], "entities": ["Moonshot AI", "Kimi K3", "OpenAI", "GLM-5.2", "DeepSeek V4", "Artificial Analysis", "Scott Galloway"], "alternates": {"html": "https://wpnews.pro/news/kimi-k3-is-not-cheap", "markdown": "https://wpnews.pro/news/kimi-k3-is-not-cheap.md", "text": "https://wpnews.pro/news/kimi-k3-is-not-cheap.txt", "jsonld": "https://wpnews.pro/news/kimi-k3-is-not-cheap.jsonld"}}