Version 4

Google launched Gemini 3.5 Flash at Google I/O 2026 on May 19, priced at $1.50/million input tokens and $9/million output tokens — 3x the cost of Gemini 3 Flash Preview [9][10]. The model replaced Gemini 3 as the default on Google's web interface, now powers Google Search AI Mode [6], and carries the tagline 'frontier intelligence with action' [1]. Google has since shifted to an enterprise cost-savings counter-narrative, with VentureBeat reporting Google's claim that the model can slash enterprise AI costs by more than $1 billion a year versus competing providers [16]. Independent benchmarks note the model scores within two points of Anthropic's flagship at a third of that flagship's price [15] — a cross-vendor frame that contrasts sharply with the developer community's experience of a 3x internal price hike.

Google's $1 billion enterprise savings claim introduces a new comparison baseline — Gemini 3.5 Flash positioned as cheap relative to GPT-5 or Claude flagship, not relative to prior Gemini Flash models. If enterprise buyers accept that frame, Google may defuse the backlash without conceding on price. The standoff between those two factually accurate but opposite-pointing comparisons will determine whether this model is remembered as an affordable frontier option or an expensive Flash replacement.

Google launched Gemini 3.5 Flash at Google I/O 2026 on May 19, marketing it under the tagline 'frontier intelligence with action' [1][2]. The model was spotted in the Google Cloud Console hours before the official announcement [3][4], alongside a companion model called Gemini Omni Flash [5]. Upon launch, Google confirmed that Gemini 3.5 Flash had replaced Gemini 3 as the default model on its web interface, that it was powering AI Mode in Google Search [6], and that it featured in updated Google AI subscription offerings announced at I/O [7]. The Gemini API changelog [8] and Google's official model blog [1] formalized its release documentation.

The pricing landed at $1.50 per million input tokens and $9 per million output tokens — 3x the cost of Gemini 3 Flash Preview and 6x Gemini 3.1 Flash-Lite [9][10]. One analysis noted the output token price is nearly 30x more expensive than the previous cheapest Flash-tier option [11]. An independent benchmark run illustrated real-world cost impact: the Artificial Analysis benchmark cost $1,551.60 on Gemini 3.5 Flash (high), compared to $892.28 for Gemini 3.1 Pro Preview [9]. These figures nearly erase the price gap between Google's mid-tier and pro offerings, undermining the cost/capability segmentation that the 'Flash' label had previously signaled to developers. Developer frustration has expanded beyond sticker shock: a Google AI developer forum thread documented unexpected cost spikes on gemini-3-flash-preview despite decreased usage, suggesting billing behavior that developers do not yet fully understand [12].

On the capability side, early signals are favorable. Artificial Analysis rated Gemini 3.5 Flash as the 'clear leader on the Intelligence vs Speed Pareto frontier' with large benchmark gains [13]. An early-access user described first impressions as feeling 'on par with Gemini 3.1 Pro Preview' [14]. A headline from RD World characterized the model as scoring within two points of Anthropic's flagship at a third of that flagship's price [15] — a cross-vendor framing that presents the model as cheap relative to Claude rather than expensive relative to prior Flash offerings. Google has pressed this alternative frame aggressively: VentureBeat reports Google's claim that Gemini 3.5 Flash can slash enterprise AI costs by more than $1 billion a year compared to competing providers [16], repositioning the pricing debate entirely away from internal comparisons.

The developer community's pricing reaction was swift and unusually intense. Posts described the pricing as 'painful' [17], announced 'the good times are over' [18], and declared outright that 'Gemini 3.5 Flash is EXPENSIVE' [19]. Reddit threads continued characterizing the price hike as a significant negative surprise [20]. At least one voice remained broadly skeptical of Google's standing despite the launch, posting that 'Google is still behind' [21]. One observer stated they had 'never seen such negative feedback in a software release before' [22]. Analyst Simon Willison contextualized the increases as part of a broader industry repricing trend — Google, OpenAI, and Anthropic all appear to be simultaneously testing API customer price tolerance, suggesting the aggressive discounting of early LLM competition may be giving way to margin recapture [9]. On the tooling side, the llm CLI plugin for Gemini added gemini-3.5-flash support in version 0.32 [23], and a concurrent alpha introduced streaming reasoning token support [24].