The Bottom Line: Google launches Gemini 3.5 Flash with immediate general availability, but integrates it into many free products—despite drastic price increases that make the model three times more expensive than its predecessor.
Google has released Gemini 3.5 Flash at its I/O conference and is rolling the flagship model directly into general availability. The catch: prices have risen substantially, while Google simultaneously integrates the model into numerous free consumer products.
Google has released Gemini 3.5 Flash today and moved the model directly into general availability without a preview phase. The company plans to integrate it into a broad range of products: the Gemini app and AI Mode in Google Search get immediate access, developers can use it via Google’s agent-first platform Antigravity as well as the Gemini API, and enterprises benefit from new enterprise agent platforms.
The new model offers a context window of over one million input tokens and 65,536 output tokens with a knowledge cutoff of January 2025. Google has also introduced a new Interactions API that mirrors OpenAI’s Responses system and enables server-side history management.
Pricing, however, marks a significant turning point. Gemini 3.5 Flash costs $1.50 per million input tokens and $9 per million output tokens—three times as much as its predecessor Gemini 3 Flash Preview and six times as much as Gemini 3.1 Flash-Lite. This approaches the pricing of the more powerful Gemini 3.1 Pro, which competes at $2 and $12 per million tokens. The upcoming Gemini 3.5 Pro is expected to be even more expensive.
A look at independent benchmarks reveals the actual burden: according to Artificial Analysis, running their benchmark test on Gemini 3.5 Flash costs $1,551.60—significantly more than Gemini 3.1 Pro Preview at $892.28. For comparison: Anthropic’s Claude Opus 4.7 and OpenAI’s GPT-5.5 show similar price increases.
Particularly noteworthy is that Google is rolling out this more expensive model simultaneously into free consumer products—a sign that all major AI labs are testing the price tolerance of their API customers.