What's this about?
Google DeepMind shipped Gemini 3.7 Flash, the latest in its fast, low-cost tier. It appeared in the newsroom feed within this digest's window and rounds out a week in which every major lab moved on price or speed rather than raw capability.
What exactly is the "Flash tier"?
Same idea as described under Qwen above: a model tuned for speed and cost rather than maximum depth. Google uses Flash for high-volume product surfaces — search summaries, in-app assistance — where latency is more valuable to the user than the last few points of quality.
What's the community saying?
There is no large dedicated discussion thread in this window. The significance is cumulative: Gemini Flash, Qwen Flash-Next, the GPT-5.6 price cut and GLM-5.3 all point at the same competitive pressure on the cheap-and-fast tier.