Back to News
ProductAI Understanding briefing

Gemini 3.8 Flash arrives in GitHub Copilot

GitHub says Gemini 3.8 Flash is rolling out across Copilot plans and coding surfaces, with introductory usage-based provider pricing through December 31, 2026.

4 min readRead the primary source
Source-provided image accompanying Gemini 3.8 Flash arrives in GitHub Copilot
Verified primary sourceFetched and verified
Publisher
github.blog
Source link
github.bloghttps://github.blog/changelog/2026-09-03-gemini-3-8-flash-is-now-available-in-github-copilot
Source type
Primary document — an official announcement, paper, filing, or first-party page we read directly.

Story last revised

ContextUnderstand this in 60 seconds

Start here

Key terms

Usage-Based Billing
Pricing where costs scale with API calls, tokens, inference time, or consumed compute.
Latency
The time between sending a request and receiving the model's output.
Test yourselfChatGPT & LLMs Quiz

What changed since publication

  1. First published
  2. This is a distinct GitHub Copilot availability announcement for Gemini 3.8 Flash, not a continuation of the archived Claude Fable 5.1 release.
  3. Republished separately after an automated matching error merged this Gemini announcement into a Claude Fable article.

What happened

GitHub announced that Gemini 3.8 Flash is available in GitHub Copilot, with a gradual rollout across supported plans and development environments.

GitHub says Google’s Gemini 3.8 Flash is now available in GitHub Copilot. It will be offered to Copilot Pro, Pro+, Max, Business, and Enterprise users, although rollout will be gradual and some users may need to check back before the model appears.

The model can be selected from the model picker in Visual Studio Code, Visual Studio Copilot CLI, GitHub Copilot cloud agent, GitHub Copilot app, JetBrains IDEs, Xcode, and Eclipse. The source does not say that every listed surface has completed rollout.

GitHub says Gemini 3.8 Flash will use introductory provider pricing under usage-based billing through December 31, 2026. It refers readers to GitHub’s model-pricing documentation but does not state the price in this announcement. Business and Enterprise administrators can manage access through Copilot settings and model policies. New models are enabled automatically under the default model-enablement setting unless an administrator has disabled the global default or this specific model.

GitHub also says that, in its early testing, Gemini 3.8 Flash performed strongly on complex terminal-based coding tasks and showed rigorous validation and persistent recovery from actionable failures. Those are GitHub’s own testing claims; the source gives no test methodology, scores, sample size, or independent verification.

Source details: github.blog

Why it matters

The announcement gives Copilot users another model option for coding tasks and extends Google’s Gemini presence across GitHub’s coding products. The practical effect depends on rollout timing, administrator settings, usage-based billing, and how the model performs in users’ own workflows. GitHub’s stated early testing is not an independent evaluation, and the source does not provide comparative scores, exact pricing, or evidence that access is immediate for every eligible user.

Adding Gemini 3.8 Flash broadens the model choices available inside a widely used coding assistant. Developers and organizations can potentially select a model suited to particular coding workflows without leaving Copilot’s existing interfaces, though the source does not establish how its capabilities compare with other available models.

The announcement is also relevant to cost and governance. Usage-based billing means actual spending may depend on requests and model use, while Business and Enterprise administrators retain policy controls. Exact pricing, quotas, regional availability, data-handling terms, and whether all features work identically across surfaces are not documented in the source.

The most consequential performance claim concerns complex terminal-based work and recovery after failures. If independently reproduced, that behavior could matter for coding agents that operate across multiple steps. GitHub’s announcement alone does not establish reliability, safety, or superiority in production.

What to watch next

Watch for the model to appear in eligible Copilot accounts, administrator enablement decisions, the exact usage-based rates, and independent testing of its terminal-based coding performance and recovery behavior.

Users should distinguish announced availability from immediate access: GitHub says the rollout is gradual, so the model may not yet appear in every eligible account or client.

Organizations should review model-policy settings before assuming employees can use Gemini 3.8 Flash. The source says new models are enabled by default only when the relevant global and model-specific controls have not been turned off.

Pricing remains a practical unknown. GitHub identifies introductory provider pricing through December 31, 2026 but does not provide rates, billing units, spending limits, or what pricing will apply afterward.

Independent evaluations should test the announced terminal-coding and failure-recovery behavior across representative repositories, including error rates, latency, cost, and the extent of human oversight required.

Related guides & quizzes

ChatGPT & LLMsAI Models ExplainedPrompt EngineeringTest what you know — try a free AI quizLook up an AI term in our glossary

Updates and corrections

This canonical story is updated in place when the developing event materially changes. Its URL and original publication date never change.

  • Republished separately after an automated matching error merged this Gemini announcement into a Claude Fable article.
  • This is a distinct GitHub Copilot availability announcement for Gemini 3.8 Flash, not a continuation of the archived Claude Fable 5.1 release.
See the public corrections log
Found this useful?