A Reshaped Google Gemini Pricing Ladder
Google is overhauling its AI subscription tiers, tightening the gap between mainstream and power-user access to Gemini. The company has introduced a new AI Ultra plan priced at USD 100 (approx. RM460) per month, positioned between its existing USD 20 (approx. RM92) and higher-end offerings. At the same time, Google has trimmed the cost of its previous USD 250 (approx. RM1,150) AI Ultra tier to USD 200 (approx. RM920), making it one of the rare premium AI packages to get cheaper rather than more expensive. This reshuffle means Google now mirrors rivals with a ladder of roughly USD 20, USD 100 and USD 200 options, while still offering a more basic AI Plus subscription around USD 8 (approx. RM37). The result is a more continuous pricing spectrum: enthusiasts, developers and small teams no longer have to jump directly from entry-level plans to the old top-end Ultra price point.
Inside the New AI Ultra Tiers and Perks
Google’s revamped Ultra lineup effectively splits the experience into a more accessible base tier and a premium high-capacity tier. The new USD 100 (approx. RM460) AI Ultra option inherits many capabilities from the former USD 250 (approx. RM1,150) plan, but with tighter usage ceilings. It offers 5x higher usage limits within the Antigravity development platform and the Gemini consumer app, plus priority access to Antigravity and 20TB of cloud storage. To bolster perceived value, Google bundles in an individual YouTube Premium subscription and YouTube Music, giving subscribers both productivity and entertainment benefits. Users on either the USD 100 or USD 200 (approx. RM920) Ultra tiers also gain early access to flagship innovations such as Gemini Spark, a 24/7 AI agent, and Gemini Omni, the company’s latest multimodal model. For those needing maximum access to cutting-edge tools like Project Genie, the refreshed USD 200 tier remains the top choice.
From Prompt Caps to Compute-Based Metering
Beyond prices, Google is fundamentally changing how it meters Gemini usage, shifting from rigid daily prompt limits to a compute-used model. Instead of counting how many prompts a user sends, Google now tracks how many tokens are consumed and how complex each request is. A simple text prompt uses far less compute than a long coding session or a rich media request, and the new system reflects that difference by charging based on actual computational load. Usage limits reset every five hours until a weekly cap is reached, smoothing out activity spikes and giving users more flexibility across the week. When subscribers hit the cap on Google’s largest models, they are automatically shifted to smaller models rather than cut off altogether. Power users on Pro and Ultra tiers can then choose to pay additional token-based API fees if they want to keep accessing the flagship models beyond their included allowance.
Strategic Implications for Consumers and Competitors
Google’s pricing and metering changes sharpen its competitive posture in the crowded AI subscription market. By aligning its USD 20, USD 100 and USD 200 (approx. RM920) tiers with similar offerings from other major labs, Google removes friction for users comparing services, while differentiating on extras like YouTube Premium and deep integration with Gemini-powered tools. Compute-based metering also better matches perceived value: light users of simple prompts avoid being penalized by blunt prompt caps, while heavy users pay more closely in line with the resources they consume. For developers and businesses, this creates clearer cost predictability and a smoother path to scaling workloads. Competitors are likely to feel pressure to match Google’s mix of lower pricing, richer bundles and more nuanced usage policies. For consumers, the net result is broader choice: entry-level, mid-tier and premium AI subscription tiers now map more cleanly to different budgets and intensity of use.