How much does Gemini 3.8 Flash cost?

Google lists Gemini 3.8 Flash at $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026, rising to $1.50 and $7.50 on 1 January 2027. A free tier is available in Google AI Studio. That introductory rate matches what Google charged for Gemini 3.7 Flash.

Who can use Gemini 3.8 Flash Cyber?

Gemini 3.8 Flash Cyber is not on general release. Google hands it to trusted defenders through the new Fairwind Program, which gives prioritised access to government authorities, critical-infrastructure operators and software maintainers. Google permits authorised threat simulation, reverse engineering and malware analysis for defensive and academic research, and forbids writing malware.

How does Gemini 3.8 Flash compare to 3.7 Flash?

Google's model card says Gemini 3.8 Flash is based on Gemini 3.7 Flash, and the launch post says it beats 3.7 Flash on the Vals Finance Agent V2 and Harvey legal agent benchmarks while keeping the same speed and price. Both models share a 1M-token context, a 64K output limit and a March 2026 knowledge cutoff.

Where can I run Gemini 3.8 Flash?

Google serves Gemini 3.8 Flash to developers through Google AI Studio, Android Studio and Google Antigravity, to companies through Gemini Enterprise, and to Google AI Pro and Ultra subscribers inside the Gemini app and Google Search. The API model ID is gemini-3.8-flash, and the minimal thinking level is not supported.

Google DeepMind · 2026-09-02 · major

Gemini 3.8 Flash — Google's best coding model yet, plus a cyber twin

Gemini 3.8 Flash is Google's new Flash-tier reasoning and coding model, built on 3.7 Flash with a 1M-token context and $0.75 per million input tokens. A separate Gemini 3.8 Flash Cyber finds and patches software bugs for vetted defenders.

Google's announcement header graphic for Gemini 3.8 Flash — Google

Google's new Flash model keeps the 3.7 price while scoring higher on coding, reasoning and security work.

Key specs

Hle verified	54.9%
Cwe bench pass@1 (cyber)	47.2%

Quick facts

Maker	Google DeepMind
Models	Gemini 3.8 Flash and 3.8 Flash Cyber
API model ID	gemini-3.8-flash
Context window	1M tokens in, 64K out
Knowledge cutoff	March 2026
Availability	AI Studio, Antigravity, Android Studio, Gemini Enterprise, Gemini app
Cyber access	Fairwind Program only (vetted defenders)

Pricing

Input · Through 31 December 2026, then $1.50	$0.75 / 1M tokens
Output · Through 31 December 2026, then $7.50	$3.75 / 1M tokens
Free tier · Free of charge in Google AI Studio	$0

source ↗

What is it?

Gemini 3.8 Flash replaces 3.7 Flash as Google's mid-tier model for coding and agent work, and Google calls it its best reasoning and coding model at Flash speed and cost. It reads text, images, audio and video across a 1 million-token window and writes up to 64K tokens back. Alongside it, Google released Gemini 3.8 Flash Cyber, a security-focused version that finds and fixes software bugs.

How does it work?

The model card says Gemini 3.8 Flash is based on Gemini 3.7 Flash rather than a fresh pretraining run, with three effort levels — low, medium (the default) and high — that trade latency against reasoning depth. On HLE-Verified it scores 54.9% across STEM, humanities and professional questions. The Cyber version is tuned for vulnerability work: Google reports a success rate above 70% at finding real bugs in codebases across 20 programming languages, and 47.2% pass@1 on CWE-Bench for writing the patch.

Why does it matter?

Cheap models doing long-horizon engineering work is the part of Gemini 3.8 Flash that changes budgets: the same $0.75 per million input tokens as 3.7 Flash now buys stronger agentic coding, so teams running agents at scale do not have to pay flagship rates. The Cyber variant is deliberately gated — Google will only hand frontier vulnerability-finding to defenders it has vetted through the new Fairwind Program, which is an admission that the same skill works for attackers.

Who is it for?

developers running coding agents, and security teams

Frequently asked questions

How much does Gemini 3.8 Flash cost?: Google lists Gemini 3.8 Flash at $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026, rising to $1.50 and $7.50 on 1 January 2027. A free tier is available in Google AI Studio. That introductory rate matches what Google charged for Gemini 3.7 Flash.
Who can use Gemini 3.8 Flash Cyber?: Gemini 3.8 Flash Cyber is not on general release. Google hands it to trusted defenders through the new Fairwind Program, which gives prioritised access to government authorities, critical-infrastructure operators and software maintainers. Google permits authorised threat simulation, reverse engineering and malware analysis for defensive and academic research, and forbids writing malware.
How does Gemini 3.8 Flash compare to 3.7 Flash?: Google's model card says Gemini 3.8 Flash is based on Gemini 3.7 Flash, and the launch post says it beats 3.7 Flash on the Vals Finance Agent V2 and Harvey legal agent benchmarks while keeping the same speed and price. Both models share a 1M-token context, a 64K output limit and a March 2026 knowledge cutoff.
Where can I run Gemini 3.8 Flash?: Google serves Gemini 3.8 Flash to developers through Google AI Studio, Android Studio and Google Antigravity, to companies through Gemini Enterprise, and to Google AI Pro and Ultra subscribers inside the Gemini app and Google Search. The API model ID is gemini-3.8-flash, and the minimal thinking level is not supported.

Try it

Select Gemini 3.8 Flash in Google AI Studio, or call the API model gemini-3.8-flash