Google launched Gemini 3.8 Flash on Wednesday, The Verge reported, three weeks after Gemini 3.7 Flash. It is the company’s third Flash model in six weeks. The model is available now to Google AI Pro and Ultra subscribers, and to developers and enterprise customers through the Gemini API, Google AI Studio, and Gemini Enterprise.
Per-token pricing matches 3.7 Flash: $0.75 per million input tokens and $3.75 per million output tokens. Google still warns that 3.8 Flash “works harder.” On complex tasks it takes extra reasoning steps and calls tools iteratively, which can mean more tokens even at the same rate. Developers who want to keep usage down can stay on 3.7 Flash, which remains supported.
Google says 3.8 Flash is stronger at software engineering and autonomous agents, and that it beats its predecessor and other frontier models on the DeepSWE v1.1 software-engineering benchmark, including Anthropic’s Fable 5. The Verge also cited wins on Vals Finance Agent V2 and Harvey’s Legal Agent benchmark. Artificial Analysis, quoted in that report, called 3.8 Flash the cheapest model it had measured at that intelligence level, even with about 30 percent more output tokens per task.
Google also released Gemini 3.8 Flash Cyber through a new Fairwind Program limited to governments and “trusted partners.” The Verge said the program has about 650 members, including CrowdStrike and the Center for Internet Security, and pairs the Cyber model with Google’s CodeMender agent to find and patch vulnerabilities. Google says the general 3.8 Flash model ships with safeguards against chemical, biological, radiological, nuclear, and cyber-offense misuse. The Cyber variant is more permissive, which is why access is gated.
Tuesday’s near-launch reports are done. The model is out.