Google Ships Gemini 3.8 Flash at $0.75/MTok with Upgraded Coding Performance

Google released Gemini 3.8 Flash on September 2, a coding-focused Flash-tier model at $0.75/$3.75 per million tokens — worth adding to your benchmark set before routing inference to it.

Google shipped Gemini 3.8 Flash (internally "Skimaki") on September 2, 2026. It's available via the Gemini API, Google AI Studio, and consumer Gemini surfaces, priced at $0.75 per million input tokens and $3.75 per million output.

Google reports internal testing in their Jetski developer platform preferred this model over Claude Opus for coding tasks — that is Google's own benchmark, not an independent evaluation, and no third-party coding benchmark results were published at launch.

Two things to verify before routing workloads to it: the knowledge cutoff varies by domain (March 2026 for some areas, January 2025 for others), which matters for tasks involving recent APIs or events. Google also shipped Gemini 3.8 Flash Cyber alongside the general model, replacing the 3.5 Cyber variant for security-focused workloads.

Why it matters: Another Flash-tier option at a price point below Sonnet-tier for longer coding runs. Run your own evals on your actual workload — Flash-tier models vary substantially by task type, and internal benchmark results from any lab rarely translate directly to production.

Sources: Gemini 3.8 Flash rolling out — 9to5Google · Google Debuts Gemini 3.8 Flash Cyber Variants — AndroidHeadlines · Gemini 3.8 Flash Launches at $0.75/1M Input — tech-insider.org