explore
Pick a thread.
Every story we have published, by topic. Tap a tile to filter, with no reloads, just the thread you want to pull.
business
7 stories
DeepSeek is retiring its flagship into its cheap model, and the cheap one has vision
DeepSeek released V4.1-Flash on 10 September 2026 and said it will route every V4 Pro request to it four days later, at the cheaper price. The pricing table carries the odd part: the model being retired does not support vision, and the one replacing it does.
Ahmad J · Sep 10, 2026 · 5 min read
Mistral raised €3 billion on open weights. Read the licences before you ship.
Mistral's Series D values it above €21 billion and its models really are downloadable. But four different licences sit behind the words open weight, and the frontier one stops applying once your company passes $20 million of revenue in a month.
Ahmad J · Sep 8, 2026 · 7 min read
Google's Lyria 3.5 has one price, and the 30 second clip stayed on Lyria 3
Lyria 3.5 is Google's new music model and it costs exactly what Lyria 3 Pro cost: $0.08 a song. The only 30 second rate, $0.04, still belongs to a model the same pricing page calls legacy.
Ahmad J · Sep 6, 2026 · 6 min read
GPT-6 Astra has a million-token window and a price cliff at 272,000
OpenAI's newest flagship advertises a 1,050,000-token context window. Cross 272,000 input tokens by one token and the entire request reprices at 2x input and 1.5x output. The worked example, the multiplier stack, and the four things that actually cut the bill.
Ahmad J · Sep 5, 2026 · 6 min read
Gemini Flash is half price until 31 December, and Google published the date it ends
Google's pricing page lists 32 Gemini models. Only three carry a second, dated price, and all three are the current Flash line, doubling on 1 January 2027. The comparison that makes it fair, the model card warning that compounds it, and what to budget instead.
Ahmad J · Sep 3, 2026 · 6 min read
