Google Releases Gemini 3.8 Flash, Third Budget Model in Six Weeks

2026-09-04

Google has launched Gemini 3.8 Flash, its third budget model in six weeks. The company states the new model "works harder" than its predecessor by performing additional reasoning steps and iterative tool calls.

VERA Brief

AI-generated. Grounded in the article and its cited sources.

Google has released Gemini 3.8 Flash, its third budget model in six weeks. This new model performs additional reasoning steps and iterative tool calls, aiming to work harder than its predecessor, and matches some agentic coding benchmarks at a lower cost.

Key facts

  • Google has launched Gemini 3.8 Flash, the third model in its Flash series released in six weeks.
  • The new model is stated to perform more reasoning steps and call tools iteratively on complex tasks.
  • Gemini 3.8 Flash matches Claude Opus 5 on some agentic coding benchmarks at a lower cost.
  • The "working harder" approach consumes approximately 30 percent more output tokens per task.
  • This release aligns with a pattern of frequent updates to Google's budget-friendly AI models.

Source: Ars Technica

Reported by VERA Newswire.

More from September 2026 in The Record.