From Cheap AI to Premium Bills: What Changed with Gemini 3.5 Flash
The end of the cheap AI era describes the shift from low-cost, subsidized models toward premium-priced AI services, where improvements in capability, speed, and reliability arrive bundled with higher subscription fees and complex pricing that favor heavy users and professional workloads over casual experimentation. Gemini’s 3.5 Flash model shows this turn clearly. According to XDA, Gemini 3.5 Flash now costs USD 1.50 (approx. RM6.90) per million input tokens and USD 9 (approx. RM41.40) per million output tokens, compared with USD 0.50 (approx. RM2.30) and USD 3 (approx. RM13.80) for Gemini 3 Flash. That is a threefold price increase on the tier meant for high-volume, low-stakes tasks. Benchmarks from Artificial Analysis found 3.5 Flash cost around 5.5 times more to run than its predecessor because it both charges more per token and tends to generate more tokens during multi-step, agent-style work.
Rising AI Subscription Costs and the New Gemini Tiers
Gemini’s subscription ladder shows how AI subscription cost is being re-framed as a gateway to premium AI features rather than pure model access. PCMag outlines four paid Google AI plans above the free tier: AI Plus at USD 8 (approx. RM36.80) per month, AI Pro at USD 20 (approx. RM92.00) per month, and two AI Ultra options priced at USD 100 (approx. RM460.00) and USD 200 (approx. RM920.00) per month. The free tier offers limited usage, while paid tiers raise those limits substantially. PCMag notes that AI Plus doubles the usage limits of the free plan, AI Pro offers four times the limits, the cheaper Ultra plan offers five times, and the top Ultra plan offers twenty times the limits. In practice, Gemini’s AI subscription cost is no longer about buying one model. It is about packaged capacity, priority, and extras that reward people who run large or frequent workloads.
Efficiency Gains Turned Into Price Increases, Not Savings
For years, the assumption was that each new generation of AI would get cheaper for a given level of intelligence. XDA reports that Gemini’s Flash line has moved in the opposite direction: Gemini 2.5 Flash launched at USD 0.30 (approx. RM1.38) per million input tokens and USD 2.50 (approx. RM11.50) for output, Gemini 3 Flash rose to USD 0.50 (approx. RM2.30) and USD 3 (approx. RM13.80), and 3.5 Flash now sits at USD 1.50 (approx. RM6.90) and USD 9 (approx. RM41.40). The hardware and research behind these models keep improving, so the raw cost per unit of intelligence is falling. What changed is who benefits: major labs have stopped passing most of the efficiency gains back to customers. As XDA puts it, they are “quietly testing how much people will actually pay,” especially for workhorse models wired into production, where switching costs are high and demand is steady.
What Premium AI Features Power Users Are Paying For
Higher Gemini pricing is easiest to understand by looking at what premium AI features paid tiers unlock. PCMag explains that all plans can access Gemini 3.5 Flash, 3.1 Flash-Lite, 3.5 Thinking, and 3.1 Pro, but the difference is how much you can use them before hitting limits. AI Plus offers moderate limits, AI Pro high limits, and Ultra plans the highest, with the top tier offering twenty times the usage of the free plan. The Pro and Ultra tiers also prioritize access to heavier models such as Gemini 3.1 Pro, which consume more of your allowance per request but are better suited for complex coding, analysis, or agentic workflows. For software developers, technical workers, content creators, and business professionals, the AI model value often lies in this combination of higher limits, priority, and access to the strongest models, rather than the headline per-token rate alone.
Bundles, Perks, and How to Decide if the New Prices Are Worth It
Gemini’s higher prices arrive alongside broader bundles that tie AI to entertainment and productivity perks. PCMag notes that paid Gemini subscriptions now sit within Google’s stack of services, where users can get AI access together with media benefits such as YouTube Premium and other account-level enhancements. This framing nudges people to see Gemini not only as an AI tool but as part of a wider digital subscription. For occasional users who mostly ask short questions, the free tier and low limits on Gemini 3.5 Flash may still be enough, even with the Gemini pricing increase in the background. For heavy users who run multi-step tasks, analysis, or high-volume content generation, choosing the right paid tier is about matching usage patterns to limits and perks. The key question is not “Is AI cheaper than last year?” but “Does this bundle provide enough AI model value for how I work, create, or build products?”
