February 2026 was honestly overwhelming for AI news, so in this video I rounded up the biggest model drops, major product updates, and the stories that actually matter if you’re building real automation systems (not just watching hype). We start with Google’s new releases: Nano Banana 2 for image generation (great quality, cheaper, a bit slower via API in my early tests), Gemini 3.1 Pro, and the underrated one that people aren’t talking about—Lyria 3 for music generation. The big takeaways here are the practical improvements: new image workflows like uploading up to 14 images, web search-assisted generation, and more “production-ready” options across Google’s stack. Then I break down Anthropic’s Claude upgrades—Opus 4.6 and Sonnet 4.6—both pushing agentic work forward with better coding, planning, and long-context reasoning, plus a 1M token context window (beta). I also cover OpenAI’s GPT-5.3 Codex (I haven’t tested it yet, but it’s clearly positioned as the top agentic coding model), Alibaba’s Qwen 3.5, and tooling updates like Cursor agents controlling VMs and Perplexity’s push toward a “computer” that can run workflows for hours. We wrap with the big industry stories: OpenAI’s $110B round at a $730B valuation, Anthropic’s Pentagon friction, and the scary side of “AI efficiency” narratives—like Block laying off nearly half its workforce.