Gemini 3.8 Flash Lands in GitHub Copilot: Which Plans Get It, and Why Yours Might Not Show It (2026)

GitHub’s official changelog on 2026-09-04 announced that Gemini 3.8 Flash is now available in GitHub Copilot. It’s a small release, but two details are worth checking before you go looking for it: whether your plan is included, and whether an admin has to switch it on.

(What follows is from reading GitHub’s official changelog on 2026-09-05.)

1. Which plans get it

The official announcement lists: Copilot Pro, Pro+, Max, Business and Enterprise.

The free plan is not on that list. If you’re on Copilot Free, this release doesn’t reach you — the model won’t appear in your picker. If you’re unsure which tier you’re on, our Copilot free vs paid breakdown covers what each tier actually includes.

2. For org accounts, the default is “on”

This is the part people get backwards. The official wording is that new models are enabled automatically unless an administrator has turned off the global default or explicitly disabled that specific model.

So:

• Individual plans (Pro / Pro+ / Max): it’s simply there. Nothing to configure.
• Business / Enterprise: usually there too. If you can’t find it in the picker, the likely cause isn’t a slow rollout — it’s that your organisation’s admin has the global default switched off. That’s a conversation with IT, not a waiting game.

On the admin side, this lives in Copilot’s model policy settings.

3. Where you can use it

The official list of supported clients runs to eight:

• Visual Studio Code
• Visual Studio
• Copilot CLI
• GitHub Copilot cloud agent
• The GitHub Copilot app
• JetBrains IDEs
• Xcode
• Eclipse

That’s fairly complete coverage — Swift developers on Xcode and JVM developers on JetBrains and Eclipse are both included.

4. One number we couldn’t verify, so we won’t print

Paid Copilot runs on premium requests, and each model carries its own multiplier, which directly determines how fast you burn your monthly allowance. This changelog does not state a premium request multiplier for Gemini 3.8 Flash.

We don’t estimate and we don’t repeat third-party guesses. To find your real consumption, check the Copilot usage page on your GitHub account after a few requests, or consult GitHub’s own model multiplier table. Don’t extrapolate from another model’s multiplier — Flash-class models are usually cheaper than flagships, but “how much cheaper” needs to come from the official table.

5. Is it worth switching to?

The Flash line has always been positioned as fast and cheap rather than smartest. The practical trade-off:

• Everyday completion, boilerplate, small fixes → Flash-class models tend to respond faster, and if the multiplier is lower you conserve allowance too. A reasonable default.
• Cross-file refactors, hard debugging, anything needing whole-architecture understanding → keep a flagship for this. Spend the allowance you saved here.

One related housekeeping note: four older models are being retired on 2 October, so don’t wire a workflow to something that’s about to leave — the list is in the four Copilot models retiring on 2 October. For how Gemini 3.8 Flash compares with the previous generation, see is Gemini 3.8 Flash worth switching to.

Checked on 2026-09-05 against the official GitHub changelog of 2026-09-04. Plan and model availability can change at any time; the official announcement is authoritative. Our plan and allowance checks live on the GitHub Copilot tool page.