Blog
Insights, tutorials, and news about LLMs, tokens, and AI tools
Google's Gemini 3.5 Flash Undercuts Everyone - Again
At I/O 2026 Google shipped Gemini 3.5 Flash: a fast, cheap model that beats its own Pro tier on coding and agentic benchmarks at $1.50 a million tokens. The 'good enough and absurdly cheap' tier just got dangerous.
Read More
Sora is Dead. OpenAI Pulled the Plug to Pour GPUs Into ChatGPT Images V2 - Here's the Real Reason
OpenAI quietly killed Sora on April 28, 2026. The official line is "product focus." The real reason is compute economics: ChatGPT Images V2 monetizes a single GPU-second far better than text-to-video ever could. Here's the math - and why every other lab is doing the same thing.
Read More
GPT-5.5 Landed at $5 a Million Tokens - and the Pricing Floor Just Moved Again
OpenAI shipped GPT-5.5 on April 23 with a million-token context window and a price that undercuts what a flagship used to cost. The benchmarks are good. The pricing is the real story.
Read More
OpenAI Images V2 Just Killed the Stock Photography and Social-Media-Design Industry
ChatGPT Images V2 dropped on April 22, 2026, and the design industry has not stopped reeling. With perfect text rendering, brand-consistent generation, and vector-quality output for $0.04 a frame, V2 leapfrogged Google's Imagen "Banana" and made Canva, Figma, and stock photography uncomfortable in the same week. Here's what changed and how creatives survive.
Read More
Google Just Put $40 Billion Into Anthropic - And Suddenly Claude's Limits Got Better. Coincidence?
On April 14, 2026, Google announced a $40 billion investment in Anthropic - a mix of cash and TPU compute credits that more than doubles Anthropic's training and inference capacity. Four days later, the long-running Claude Pro limits drama eased. We can't prove causation, but the timeline is impossible to ignore. Here's what changed and what it signals.
Read More