Google released Gemini 3.6 Flash at lower prices than its predecessor while launching two additional models. The new flagship costs $1.50 per million input tokens and $7.50 per million output tokens.
Google announced three new Gemini models Tuesday, reshaping its AI pricing structure with more aggressive positioning against competitors.
Gemini 3.6 Flash undercuts the previous generation with pricing at $1.50/1M input tokens and $7.50/1M output tokens—a notable reduction from 3.5 Flash's rates. The model delivers improved performance on coding, knowledge work, and multimodal tasks while reducing output token usage by up to 17% versus 3.5 Flash.
Gemini 3.5 Flash-Lite targets cost-sensitive applications at $0.30/1M input tokens and $2.50/1M output tokens, positioning itself as Google's most affordable option for lighter workloads.
Google also introduced Gemini 3.5 Flash Cyber, a specialized model addressing cybersecurity use cases where Anthropic has established early market presence.
The pricing moves reflect intensifying competition in the AI model market, where cost efficiency has become a primary differentiator. By reducing prices on improved models, Google aims to accelerate adoption among developers building AI agents and applications requiring high throughput at scale.
Looking ahead, Google confirmed it has begun pre-training for Gemini 4, described as its "most ambitious pre-training run yet." The company also teased an upcoming Gemini 3.5 Pro, though no release date or pricing has been announced.
These announcements arrive as major AI labs compete on both capability and economics. Google's strategy prioritizes efficiency and cost reduction alongside performance improvements, signaling confidence in its ability to deliver stronger models at lower price points. The expanded model lineup—ranging from specialized cybersecurity variants to ultra-low-cost options—allows Google to serve different customer segments simultaneously.
Developers can access the new models through Google's AI API and cloud services, with Gemini 3.6 Flash available immediately.
The LA Unified School District has barred approximately 378,000 students from accessing generative AI tools on district-provided laptops and tablets. The ban takes effect while officials conduct a comprehensive review of AI's role in educational settings.
WebLLM is a high-performance inference engine that runs large language models directly in web browsers without server dependency. The open-source project from MLC AI enables client-side LLM execution with optimized performance.
Meta is reducing mandatory AI tool usage requirements for employees while simultaneously promoting Hatch, its latest advanced AI agent designed for internal use.
Meta has released Muse Spark 1.3, an updated AI model available to developers. The release garnered significant developer interest on Hacker News with 167 points and 89 comments.