Two stories from the last few days capture the strange state of the AI industry heading into the back half of 2026: everyone is shipping faster than ever, yet the model everyone actually wants is still nowhere to be found.

Flash over substance

On August 13, Google released Gemini 3.7 Flash, its third Flash-tier update in a matter of months. Google has released Gemini 3.7 Flash, the newest model in its Flash tier, three weeks after Gemini 3.6 Flash, with the model card describing it as a refinement of 3.6 Flash with algorithmic improvements to the core reasoning foundation rather than a new pretraining run. The pitch is squarely aimed at developers: it accepts text, images, audio, and video across a 1M-token context window, returns up to 64K output tokens, and supports customizable thinking configurations that trade quality against cost and latency.

What's most notable isn't the benchmark bump — though the DeepSWE v1.1 benchmark goes from 49.0% to 65.3%, while FrontierCode 1.1 Main climbs from 34.4% to 43.6%, and WebDev Arena Elo rose from 1538 to 1588 — it's the price. Gemini 3.7 Flash ships at $0.75 per 1M input tokens and $3.75 per 1M output tokens, half the original 3.6 Flash list rate and roughly a third the blended cost of Claude Sonnet 5 or GPT-5.6 Terra. That discount isn't permanent, though: starting January 1, 2027, pricing rises to $1.50 per million input tokens and $7.50 per million output tokens.

Where's the flagship?

The subtext to every Flash release is Google's absent flagship. Alphabet released another installment of its workhorse Gemini Flash model but didn't say when it would release its most powerful AI model, Gemini 3.5 Pro, which has been hampered by delays. A Forbes report published the same day as the Flash launch notes the model has become the "longest-awaited model of 2026," with reports citing persistent coding and reliability problems, senior researcher departures, and a possible complete retraining from its foundational pre-training phase due to a "structural problem." Originally teased for a June launch at I/O, the Pro model has now slipped by months, and rival labs including Anthropic, OpenAI, and Moonshot have used the gap to push out their own coding-focused releases.

A speed-and-pricing war, not a hype cycle

Zoom out and the pattern repeats across the industry. In the first two weeks of August alone, trackers logged roughly ten new model releases from six providers — ByteDance's Seed 2.1 Turbo, xAI's Grok Imagine Image 2.0, and Google's own Gemini 3.7 Flash among them — with new flagship-adjacent models arriving on an almost weekly cadence. Rather than one breakthrough dominating headlines, the story is a market where labs compete on cost-per-token and turnaround time as much as raw capability, squeezing margins across the board.

Meanwhile, that intensity is showing up in markets too. Investment analysts noted this week that this earnings season reinforced a theme that has driven equity markets for quarters: AI remains one of the most important sources of growth across the economy, with cloud demand continuing to accelerate and prompting many of the largest tech platforms to raise capital spending expectations again. Google's stumble on Gemini 3.5 Pro is a reminder that even the deepest-pocketed labs aren't immune to the pressure that spending is creating.

Sources

AI IndustryAI ModelsDaily AI UpdateGeminiGoogle DeepMindLarge Language Models