Google launched Gemini 3.8 Flash, its latest Flash-tier model, and positioned it for long-horizon software engineering and autonomous agent tasks.
The model is being offered at an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens, but the company cautioned the model might use more tokens to maximize performance especially at higher effort levels.
Aigora.ai CEO John Ennis called the model "Opus 5 coding quality but at a fraction of the cost and super fast," adding that "This is going to be so awesome for things like making remotion videos."
Iterative call
Google says Gemini 3.8 Flash performs more reasoning steps and "calls tools iteratively" to tackle complex problems calling tools iteratively.
Google also highlighted benchmark gains, saying 3.8 Flash outperforms Gemini 3.7 Flash and other frontier models on the DeepSWE v1.1 software engineering benchmark and on the Vals Finance Agent V2 and Harvey's Legal Agent benchmarks.
Early feedback
Early third-party testing flagged higher per-task output and cost: Artificial Analysis reported about a 30% rise in output tokens and an around 40% cost increase versus 3.7 Flash in agentic evaluations 30% increase in output tokens per task.
Google is offering developers the option to continue using Gemini 3.7 Flash if minimizing token usage is a priority keep using Gemini 3.7 Flash, and it is launching a cyber-specialized variant, Gemini 3.8 Flash Cyber, into its Fairwind Program.
