Google is preparing a new AI model, Gemini 3.8 Flash, after internal tests indicate progress in coding-related tasks, an area where the company has previously been behind competitors. The Wall Street Journal reports that the latest results come from internal evaluations rather than an announced public benchmark.
Multiple outlets frame the development as an effort to narrow the gap in coding ability versus rival systems from firms such as Anthropic and OpenAI. The Wall Street Journal adds that a release is expected this week, suggesting the company is moving quickly from internal testing toward rollout. Investing.com similarly reports that Google is preparing Gemini 3.8 Flash to address the perceived coding-performance shortfall.
While both sources align on the purpose and timing, they differ in how much technical detail they provide. Neither source specifies benchmark scores or test methodologies in the provided excerpts, and both rely on reporting of internal performance rather than publicly verifiable results.
