Gemini 3.8 Flash

Available

Google's stable Flash model for long-horizon software engineering, autonomous agents, and complex enterprise workflows.

Access details: Gemini API

The model ID to use in API requests.

gemini-3.8-flash

Updated Oct 20, 2026

Sources & notes

Capabilities & limits

Context window: 1M
Maximum output: 64K
Image input: Supported
Tool calling: Supported
Structured output: Supported

Pricing

USD / 1M tokens

Input: $0.75
Output: $3.75

Standard rates per million tokens. Discounts and long-context rates, where offered, are on Google's pricing page.

Full rate card 8 rates

All prices in USD.

Standard input introductory

$0.75 USD per 1M tokens

effective through 2026-12-31

Standard output introductory

$3.75 USD per 1M tokens

includes thinking tokens
effective through 2026-12-31

Standard cache read introductory

$0.075 USD per 1M tokens

effective through 2026-12-31

Cache storage introductory

$0.50 USD per 1M tokens per hour

effective through 2026-12-31

Standard input

$1.50 USD per 1M tokens

effective from 2027-01-01

Standard output

$7.50 USD per 1M tokens

includes thinking tokens
effective from 2027-01-01

Standard cache read

$0.15 USD per 1M tokens

effective from 2027-01-01

Cache storage

$1.00 USD per 1M tokens per hour

effective from 2027-01-01

Free tier is free of charge subject to availability and limits; Batch, Flex, and Priority tiers are separately priced.

Price history

Sep 2, 2026

Input $0.75 per million tokens

Jan 1, 2027

Input $1.50 per million tokens

Sep 2, 2026

Input $0.75 per million tokens

Sep 2, 2026

Output $3.75 per million tokens

Jan 1, 2027

Output $7.50 per million tokens

Sep 2, 2026

Output $3.75 per million tokens

Technical details

Parameters: Not published
Active parameters: Not published
Architecture: Based on Gemini 3.7 Flash
Training / knowledge cutoff: March 2026
Input modalities: Text, image, video, audio, PDF
Output modalities: Text

From Google

Official announcement ↗

How they describe it

Our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.

What they say is new

delivering significant improvements from 3.7 Flash across software engineering, agentic tasks, and critical, multi-step reasoning in specialized domains

Release & lifecycle

First available: Sep 2, 2026
available: As of Sep 13, 2026
available: As of Oct 20, 2026