On July 16, Moonshot launched Kimi K3, a 2.8 trillion parameter model and the largest open model ever released. Demand was so intense that Moonshot's first-party API sold out within 48 hours. Then this weekend the full open weights went live, all 594 gigabytes of them,
On July 24, Jensen Huang did something he had never done before. The CEO of the world's most valuable company made his first post on X. It was not a product announcement or a victory lap. It was a three page policy letter to Washington titled Open Weights
On July 24, Anthropic shipped Claude Opus 5. It lands within 0.5% of Fable 5's peak intelligence at half the cost per task, priced at $5 per million input tokens and $25 per million output tokens, the same as Opus 4.8. It set a new state
On July 22, the White House got into the business of model training pipelines. OSTP chief Kratsios publicly accused Moonshot AI of distilling Anthropic's Fable model without authorization to build Kimi K3. Treasury Secretary Scott Bessent said sanctions remain on the table. The Hill added a second charge,
Every pitch deck I've seen this year has the word autonomous in it. Usually in the first three slides, usually in a font size that suggests the founder believes it. The agentic AI market is projected to grow from $7.8 billion to $52 billion by 2030. Gartner
On July 16, Moonshot released Kimi K3, a 2.8 trillion parameter open-weight model, the largest ever put out in the open, benchmarked directly against the top closed systems in the US. Forty-eight hours later, Moonshot paused new subscriptions because demand had pushed their GPU capacity to its
The self-driving company is the pitch of the summer. Forbes profiled it on July 20 as AI's next business model, pointing to Replit, where agents investigate production incidents, answer product questions, analyze business data, enrich sales leads, build presentations, and triage support tickets. MIT Sloan says these
Everyone in AI is staring at the wrong chart. The benchmark leaderboards get the headlines, the model releases get the launch videos, and meanwhile the chart that actually determines which startups survive is quietly going vertical. Inference cost for GPT-4-class capability fell from $30 per million tokens in
Last week, on July 15, Thinking Machines Lab shipped its first model. Mira Murati's lab called it Inkling, and on paper it looks like another big release: 975 billion parameters in a Mixture-of-Experts design, 41 billion active per token, multimodal, Apache 2.0 open weights. But
Last week settled an argument I have been having with people for two years. On July 15, Thinking Machines Lab, Mira Murati's outfit staffed heavily with ex-OpenAI leadership, shipped its first model. Not a demo, not a waitlist, not an API with usage tiers. Inkling landed as
Let me state the tradeoff honestly, because most people arguing my side won't. Open-source models lag the closed frontier. The gap is real, roughly three to six months depending on the task. If you rent frontier capability from a closed API, you get the best model on
There is a slide deck making the rounds in 2026, and if you have sat through one enterprise AI pitch this year, you have seen it. Map your workflows. Deploy agents to automate them. Collect the savings. Every consultancy has a version. Every agent platform has a version. The logos