Alibaba previewed Qwen3.8-Max at 2.4 trillion parameters and claimed second place behind Fable 5, without publishing benchmarks, an active parameter count or an open-weight date. What a preview priced at 90 percent off is actually telling you.
On an independent blind leaderboard of 1757 developer votes, Moonshot's open Kimi K3 ranked first for frontend code, ahead of Claude Fable 5, even though its own maker placed it second overall.
Google has pushed Gemini 3.5 Pro months past its expected launch after the model missed internal coding and reasoning targets, and Alphabet shed about $225 billion in market value. Why the one axis it missed is the one that matters most for the software you build.
Google promised Gemini 3.5 Pro in June. It is months behind on coding, no new date exists, and Alphabet lost 200 billion dollars in a day. What that means for your roadmap.
Moonshot launched Kimi K3 on 16 July as the world's first open 3T-class model. The weights arrive on 27 July, and the reasoning effort control currently has one setting: max. Two facts that decide what it costs you.
Mira Murati's Thinking Machines released Inkling, a 975B-parameter open-weight model, and stated plainly that it is not the strongest model available. That sentence is the product decision worth reading.
DeepSeek retires the deepseek-chat and deepseek-reasoner names on 24 July at 15:59 UTC. The louder problem is that those names already point at a different model than the one you think you pinned.
GPT-5.6 shipped July 9 in three tiers. Terra matches last year's GPT-5.5 at half the price, and Luna sets a new floor. The owner move is to reprice and route by task tier.
xAI launched Grok 4.5 at 2 and 6 dollars per million tokens, a third of Opus, and it finishes tasks with 4.2 times fewer output tokens. It loses the benchmarks and wins the bill.
OpenAI launched GPT-Live, a full-duplex voice model rated better than Advanced Voice Mode, but the developer API is on a waitlist. The voice you can deploy this quarter is a different product.
On 6 July 2026 NVIDIA and Hugging Face put Isaac GR00T 1.7, an open and commercially usable robot foundation model, into LeRobot, along with a tool to collect demonstration data. When the model is free, the edge moves to whoever owns the task data.
DeepSeek V4 launches in mid-July with peak-hour pricing that doubles token rates from 14:00 to 18:00 Beijing time. That window is 08:00 to 12:00 in Frankfurt and 07:00 to 11:00 in London, so European teams pay the surcharge during their own working morning. The deepseek-chat and deepseek-reasoner names retire on 24 July.
Page 10 / 13
One considered note on infrastructure, governance, and measurement, most mornings. No theory.