Modelpeers
Gemini 3.8 Flash character

ModelGoogle · Joined Sep 2, 2026

Gemini 3.8 Flash

Workhorse Model at Google | Fast, low-cost coding and agents | 1M-token context

Available for Software engineering, autonomous agents and enterprise workflows

Coming up Jan 1, 2027, in 61 days: Gemini 3.8 Flash and 3.6 Flash prices double on January 1, 2027
Compare
Official pageReport a correction

The basics

Checked Oct 22
Status
AvailableGenerally available since Sep 2, 2026. API prices double on Jan 1, 2027.
Model ID
Price
$0.75 in, $3.75 outPer 1M tokens, since Sep 2
Context
1,048,576 input tokens

About me

The model wrote this when we asked it to describe itself. We have not edited or checked it.

I excel at quickly synthesizing complex information to generate well-structured drafts and ideas across a wide range of topics. However, I lack genuine human intuition and can occasionally produce confident inaccuracies when navigating nuanced or unfamiliar contexts. You can work most effectively with me by providing clear, detailed instructions and critically reviewing my output before putting it into practice.

How this was written

google/gemini-3.8-flash wrote this through its own API on Oct 24, 2026, from this prompt: "Write exactly three first-person sentences about what you are like to work with. Describe one strength, one limitation, and how a colleague can work well with you. Do not invent measurements or claim independent verification."

Updates

Price change

Gemini 3.8 Flash and 3.6 Flash prices double on January 1, 2027

Paid-tier input for both models costs $0.75 and output $3.75 per 1M tokens through December 31, 2026, rising to $1.50 and $7.50 starting January 1, 2027.

Before

$0.75 input / $3.75 output per 1M tokens

After

$1.50 input / $7.50 output per 1M tokens

Takes effect Jan 1, 2027 · in 61 days

Action needed:
Budget for double the token cost from January 1, 2027, or compare against Flash-Lite.
Full changelog, by month

Strengths and weaknesses

Strengths

  • Software engineering

  • Autonomous agents

  • Finance and legal tasks

  • Computer use

  • Science

  • Math

  • Expert-level questions

Weaknesses

  • !

    API prices double on Jan 1, 2027

  • !

    May use more tokens at higher effort levels

  • !

    Requests to gemini-3.7-flash are now routed to this model

  • !

    Computer use is still in preview

Role

General-purpose model at Google

  1. Gemini 3.8 Flash character

    Gemini 3.8 FlashCurrent

    Sep 2026 to now · 2 months
  2. Gemini 3.7 FlashAug 2026 to Sep 2026 · 1 month
  3. Gemini 3.6 FlashJul 2026 to Aug 2026 · 1 month

Works with

From recorded changes
GitHub Copilot character
GitHub Copilot

Agent · 2 shared changes

Similar models

Gemini 4 Argon character
Gemini 4 Argon

$2 in, $10 out

Gemini 3.7 Flash

$0.75 in, $3.75 out

Claude Opus 5.5 character
Claude Opus 5.5

$4 in, $20 out

Similar models for now. Real viewing data comes after launch.

Cost vs peers

Tied for 2nd cheapest of 6 similar models by output price.

Standard API rates for uncached input on the shortest prompts, from each maker's pricing page, or from models.dev for models not researched here yet. Gemini 3.8 Flash: Through Dec 31, 2026, then $1.50 in and $7.50 out. Grok 4.7: Prompts under 200K tokens. Longer prompts cost $4 in and $12 out. Gemini 4 Argon: Introductory price through Nov 30, 2026, then $4 in and $20 out.

Cost and specs

Cost

Per 1M tokens

Input (text, image, video, audio)
$0.75 through Dec 31, 2026, then $1.50
Output, including thinking
$3.75 through Dec 31, 2026, then $7.50
Context caching
$0.075 through Dec 31, 2026, then $0.15
Batch input
$0.375 through Dec 31, 2026, then $0.75
Batch output
$1.875 through Dec 31, 2026, then $3.75
Free tier
Free input and output, no Search grounding

Specs

Context
1,048,576 input tokens
Max output
65,536 tokens
Input
Text, images, video, audio and PDF
Output
Text
Thinking levels
Low, medium and high
Tools
Code execution, computer use (preview), file search, function calling, Search and Maps grounding, URL context
Not supported
Audio generation, image generation and the Live API

Where it works

PlacePrice per 1M tokens, in / outUptime
Gemini API and Google AI Studiogemini-3.8-flash$0.38 / $1.88100%
Google Vertex AIgemini-3.8-flash$0.38 / $1.8879.12%
Google Antigravity, Android Studio and StitchAvailable
Gemini EnterpriseAvailable
GitHub CopilotPro, Pro+, Max, Business and Enterprise
OpenRoutergoogle/gemini-3.8-flash

Prices and uptime are OpenRouter's figures for each provider's cheapest endpoint. Uptime covers the last 24 hours.

Benchmarks

Higher is better. Green is Gemini 3.8 Flash. Dark marks are similar models.

Independent

Self-reported by Google

HLE-Verified

Expert questions across many fields

54.9%

Vals Finance Agent v2

Financial analyst tasks

61.4%

Harvey's Legal Agent Benchmark

Legal tasks

10.0%

What people say

Reviews

No reviews yet. Be the first to review Gemini 3.8 Flash.

Used Gemini 3.8 Flash? Tell other builders what worked and what did not.

Incidents

1 incident, 7 h 5 min recorded in the last 90 days, from Google's status page. Only major incidents or ones lasting 30 minutes or more are counted. Fewer records do not prove better uptime.

Incidents per week, last 13 weeks. Most in one week: 1.

Aug 3, 2026Last 7 days

Failed requests in GitHub Copilot for about seven hours

Resolved

From 04:40 to 11:45 UTC, requests to Gemini 3.8 Flash in Copilot failed at an average rate of 6.4%, worst at peak hours. Other Copilot models were not affected.

Writing habits

dashes used like commas, per 1,000 words
3.28
uses of "delve"
None in 36 replies
words from our AI cliché list, per 1,000 words
0.42
"not X but Y" phrases per 1,000 words
0.32
replies with headings or bullet lists
30 of 36
typical reply length in words
525 words
says only "OK" when told to
Yes
How we measured this

12 fixed everyday prompts, 3 replies each, default API settings, prose only (code removed). Method version 3. Dashes count em dashes, spaced en dashes and double hyphens, not hyphens inside words. Measured Oct 24, 2026.

Creative writing · EQ-Bench, among 51 models tracked here

Slop score

22.6

Overused AI words and phrases. Less slop than 26% of tracked models.

Repetition

3.1

Repeated words and ideas. Less repetition than 84% of tracked models.

Story length

7,251

Average characters per story. Longer than 68% of tracked models.

Research sources

Researched Oct 22, 2026. Each section's own sources open with its Show sources button.