GitHub Copilot adds Gemini 3.8 Flash with usage-based billing
GitHub is rolling Google’s Gemini 3.8 Flash into Copilot for Pro, Pro+, Max, Business, and Enterprise users, with introductory provider pricing under usage-based billing through December 31, 2026.
GitHub is adding Google’s Gemini 3.8 Flash to GitHub Copilot for Pro, Pro+, Max, Business, and Enterprise users. The rollout is gradual, and GitHub says the model will use introductory provider pricing under usage-based billing through December 31, 2026.
That makes this a Copilot availability change, not a new Gemini launch. Independent model analysis lists Gemini 3.8 Flash at $0.75 per million input tokens and $3.75 per million output tokens, with a one-million-token context window; GitHub’s September 4 notice extends the model into another widely used developer surface.
Gemini 3.8 Flash enters the Copilot model picker
Users will be able to select Gemini 3.8 Flash from the Copilot model picker as the rollout reaches their account. GitHub does not promise immediate access for every eligible user, so a missing model is currently consistent with the stated staged deployment.
The model is aimed at complex terminal-based coding tasks, multi-step reasoning, and agent workflows. GitHub describes early testing with strong validation and recovery from actionable failures. Those are GitHub’s observations, not a guarantee that the model will outperform a team’s current choice on its own repositories.
Plans, billing, and administrator controls
Gemini 3.8 Flash is listed for Copilot Pro, Pro+, Max, Business, and Enterprise. The introductory pricing is attached to usage-based billing, so teams should verify how model requests and included allowances are accounted for in their plan before enabling it broadly.
Business and Enterprise administrators can control access through Copilot model policies. GitHub says new models are enabled automatically under the default model setting unless an administrator has disabled the global default or explicitly turned off this model. Organizations with strict model allowlists should therefore check policy behavior during the rollout.
The stated pricing window runs through December 31, 2026. GitHub’s notice does not establish what pricing or billing treatment will apply afterward; that remains a future planning point rather than a promise of continued introductory rates.
What Gemini 3.8 Flash changes for coding agents
Independent analysis places Gemini 3.8 Flash among the stronger models in its price range and records support for text, image, speech, and video input with text output. Those measurements describe the underlying model, not guaranteed Copilot behavior. The separate Gemini 3.8 Flash Cyber variant is a restricted security offering and is distinct from the Copilot model listing.
For teams evaluating the Copilot rollout, the useful checks are concrete: confirm the model appears in the picker, record whether requests consume included or metered usage, compare tool-call recovery on representative tasks, and test repository instructions and policy controls. Keep the model disabled for sensitive repositories until data-handling and administrator settings match the organization’s requirements.
GitHub’s next practical milestone is wider rollout across eligible plans. Until the picker and billing details are visible in a specific account, treat availability as staged rather than universal.
Sources
Try the related loot
Put six hosted Workers AI models behind Cloudflare AI Search