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 Ships Gemini 3.8 Flash, Its Third Flash Release in Six Weeks

While Pro updates stall, Google keeps iterating rapidly on its lighter-weight Gemini model.

Google has quietly rolled out Gemini 3.8 Flash, marking the third Flash-tier model release in just six weeks. Flash models are designed to be faster and cheaper than Google's flagship Pro models, trading some raw capability for speed and lower cost, making them popular for high-volume or latency-sensitive applications.

Notably, updates to Gemini's Pro line appear to have stalled, with Google instead concentrating its visible development effort on the Flash series. Details on what specifically changed in 3.8 versus its immediate predecessors remain thin, suggesting these may be incremental tuning updates rather than major architectural shifts.

The rapid cadence signals Google is treating Flash as its primary testbed for quick iteration, possibly using it to experiment with training techniques or respond to competitive pressure before committing changes to the more resource-intensive Pro tier.

Why it matters: For developers building on Gemini's API, this pace means benchmarks and prompt behavior can shift every couple of weeks, so pinning model versions in production is more important than ever. It also hints that Google may be prioritizing cost-efficient inference over frontier capability gains for now.

Sources: Ars Technica