GKRootWire
Security 153 Million Driver's Licenses Reportedly Leaked in Massive BreachSecurity Palo Alto Networks Reportedly Pays $500M for AI IT Automation Startup ConsoleDev Tools Wasmi v2.0 Pushes WebAssembly Interpreters to New Speed LimitsDev Tools How to Reverse Engineer Unknown File Formats with ImHexDev Tools The Original Microsoft Source Code: Altair BASIC From 1975 ResurfacesSecurity Attackers Exploit Sangoma Switchvox Bug to Plant Reverse ShellsSecurity 153 Million Driver's Licenses Reportedly Leaked in Massive BreachSecurity Palo Alto Networks Reportedly Pays $500M for AI IT Automation Startup ConsoleDev Tools Wasmi v2.0 Pushes WebAssembly Interpreters to New Speed LimitsDev Tools How to Reverse Engineer Unknown File Formats with ImHexDev Tools The Original Microsoft Source Code: Altair BASIC From 1975 ResurfacesSecurity Attackers Exploit Sangoma Switchvox Bug to Plant Reverse Shells
AI

Google's New Gemini 3.8 Flash Thinks Harder, But That Thinking Isn't Free

Google's latest Flash model adds extra reasoning steps and iterative tool calls, which could quietly inflate your API bill despite unchanged sticker pricing.

Google has rolled out Gemini 3.8 Flash, replacing 3.7 Flash just weeks after that model's debut. The headline change isn't a price cut or a benchmark flex — it's that the model is designed to "work harder" by running more reasoning steps and calling external tools repeatedly when tackling complex tasks.

Pricing stays the same as 3.7 Flash: $0.75 per million input tokens and $3.75 per million output tokens. But Google itself cautions that the model may burn through more tokens to hit peak performance, particularly at higher "effort" settings, meaning actual costs could climb even though the rate card didn't change. Developers who want predictable, lower token usage can stick with 3.7 Flash instead.

The release has already drawn some early scrutiny from developers testing the tradeoffs in practice.

Why it matters: This is a preview of how AI vendors may increasingly compete: not on transparent per-token pricing, but on opaque 'effort levels' that shift real costs onto usage rather than list price. Teams building on Gemini should budget for token volume, not just rate cards, and treat model upgrades as cost experiments rather than free wins.

Sources: The Verge