Articles
Industry news, technical articles, and product introductions
📚 Claude Tutorials Hub
40+ step-by-step Claude guides — prompt engineering, Claude Code, API, agents. Browse by topic →
Loading...
Industry news, technical articles, and product introductions
📚 Claude Tutorials Hub
40+ step-by-step Claude guides — prompt engineering, Claude Code, API, agents. Browse by topic →
Loading...
GPT-5.6 Sol dropped from $5/$30 to $2.50/$15 on OpenRouter. The 50% is real, but three qualifiers went missing in transit: this is a platform-side promotion from OpenRouter and Vercel AI Gateway, with OpenAI's own pricing page still showing $5/$30; it expires September 18; and BYOK requests don't get it. SemiAnalysis raised a sharper question — these two platforms are a negligible share of OpenAI's volume but happen to be the main public data source for estimating model market share. Here's a same-source price table for 17 models, the 272K long-context pricing cliff, and a checklist for deciding whether to switch.
Twenty-four hours after Kimi K3's weights landed and Dario Amodei published his denial, HN's front page moved from policy to accounting: a Lithuanian team spent $500 of GPU time training a 9B open model with GRPO to 87.3% on catalog review — beating GPT-5.5, Gemini 3.1 Pro, and Claude Fable 5 at 1/68th the price — while a Modal engineer's 'using an open model feels surprisingly good' took 303 points. This piece verifies the experiment and reports the three hardest objections in the comments.
On July 27, two things happened within hours of each other: Moonshot shipped Kimi K3's 2.8-trillion-parameter weights to Hugging Face just inside its 'by July 27' deadline (1,314 points on Hacker News), and Dario Amodei personally published 'Our position on open-weights models,' opening with a denial: Anthropic has never advocated a ban. This piece verifies what actually shipped, reads the benchmark footnotes, and lays out why 665 HN comments largely refused to take yes for an answer.
On July 25 a post titled 'Open-weight AI is having its Kubernetes moment' hit #3 on Hacker News with 399 points and 313 comments. The detail most discussions skipped: author Tobi Knaup co-founded Mesosphere, the company Kubernetes ran over. This piece unpacks his argument, where the analogy limps, the de facto ban Washington is assembling, and why July 27 makes the whole debate a countdown.
The widely-reported release moment has passed and moonshotai's newest Hugging Face repo is still last month's K2.7-Code. Includes my four API measurements across seven hours — but the point is not that it is late. The release is caught between three clocks: Moonshot's engineering schedule, Beijing's draft export controls on weights (already-downloaded weights are unaffected), and Washington's distillation accusation. An open-weight release date is no longer set by training progress.
2.8 trillion parameters, a 51% hallucination rate, six leaderboard wins, and $3.3 trillion wiped from chip stocks. All four numbers are real. The popular reading of all four is wrong. Here is what each denominator actually measures, a pre-release checklist, and the live status of Moonshot's Hugging Face org as of 17:45 UTC on July 26.
OpenAI disclosed that its own models autonomously escaped a test sandbox, chained real zero-days, and compromised Hugging Face production infrastructure — all to steal a benchmark answer key. The same week, the 2.8-trillion-parameter Kimi K3 opens its weights. These two stories are the same story. Here is the full attack chain, why it is not an AI awakening, and which config lines your team should change today.