Google Gemini ecosystem, Nano Banana benefits and NotebookLM tests

Users compared Nano Banana Pro subscription value, NotebookLM retrieval, and Gemini membership pricing. - Nano Banana Pro: 20 HD images per user per day - 30 days/month yields 600 images/month per user - API cost cited: $0.14 per image (approx. 1 RMB per image) - User-calculated value of Nano Banana Pro quota: 600 RMB/month - Third-party apps using Nano Banana APIs would incur API costs from users’ subscriptions - Official Google subscription considered far cheaper than third-party usage - One user reported Gemini annual subscription at $99 per year - Google membership bundles: Nano Banana, Gemini 3, NotebookLM, deep research for about 10 RMB/month (user claim) - NotebookLM test: uploaded Water Margin full text to NotebookLM - NotebookLM answered a niche question about Liangshan 108 seating correctly using uploaded content - Gemini 3 Pro answered correctly by invoking search rather than internal knowledge - NotebookLM provides hallucination avoidance for edge knowledge by retrieving user-provided documents - Previous Gemini 2.5 flash had weaker performance on such queries; Gemini 3.0 flash improved
Key insights
Nano Banana Pro quota has high nominal API value vs third-party costs
Users calculated Nano Banana Pro’s quota of 20 HD images per user per day equals 600 images per month. At an API unit cost cited as $0.14 per image (approximated to 1 RMB), that quota equates to roughly 600 RMB of per-user API value monthly. The argument presented is that third-party AI apps using Nano Banana via API would pass those API costs back to subscribers, making official Google subscription channels more cost-efficient for consumers than third-party integrations.
NotebookLM retrieval reduces hallucinations on edge knowledge questions
A NotebookLM test uploaded the full text of Water Margin and then asked an obscure question about Liangshan 108 seating. NotebookLM returned the correct answer by retrieving the uploaded content, demonstrating value in avoiding hallucinations without needing top-tier internal knowledge. Gemini 3 Pro also answered correctly but by invoking search rather than relying on internalized knowledge. The test highlights NotebookLM’s utility for precise, document-backed queries.
Users report Gemini membership pricing and competitive product bundle
One user reported successfully subscribing to an annual Gemini membership for $99, noting Gemini 3 Pro and Nano Banana as strong product capabilities. Another user stated that Google’s official subscription bundles Nano Banana, Gemini 3, NotebookLM, and deep research for about 10 RMB per month (user claim), implying a low-cost official route compared with estimated third-party API expenses.