GitHub announced Gemini 4 Hydrogen for Copilot on November 13, 2026, giving developers and software teams another model choice inside their existing coding tools. The company calls it generally available, but says the rollout is gradual.
Who can select it
The announcement lists Copilot Pro, Pro+, Max, Business and Enterprise. Once available to an account, Hydrogen can be selected in the model picker across supported products, including VS Code, Visual Studio, JetBrains, Copilot CLI, the coding agent, github.com and GitHub Mobile. GitHub has not given a rollout completion date.
There is an important qualification to that paid-plan list. GitHub’s legacy billing documentation says Pro and Pro+ users who kept annual request-based billing after June 1, 2026 do not receive new models or features. Check that billing arrangement before treating a missing Hydrogen option as a rollout delay.
On workplace accounts, administrators control model access. New models are enabled by default unless an administrator has changed that default or disabled the model. An enterprise-enforced restriction cannot be overridden by an organization owner.
Copilot Free and Student offer automatic model selection only, rather than the manual model choice described here.
What Hydrogen usage costs
GitHub’s current rate card lists Hydrogen at US$1.50 per million input tokens, $0.075 per million cached input tokens and $7.50 per million output tokens. Tokens are the units used to count what the model processes and generates; cached input is reused context. One AI credit equals $0.01.
The subscription supplies an allowance, not a fixed number of prompts. Current monthly individual plans include 1,500 credits on Pro ($10), 7,000 on Pro+ ($39) and 20,000 on Max ($100). These totals include a variable flex allotment, so check the account’s current allowance.
Business and Enterprise licenses contribute 1,900 and 3,900 credits respectively each month to a shared pool, rather than separate personal balances. Usage draws from that pool before additional paid usage.
Consuming included credits is not itself an extra invoice charge. Nor does the token tariff establish a price per completed coding task: one agent task can involve several model calls, with costs depending on how much each processes and generates.
Set access and spending separately
For teams, allowing Hydrogen and allowing extra spending are separate decisions. Business and Enterprise paid usage is enabled by default; to prevent spending beyond the included pool, an administrator must disable the AI credits paid usage policy.
For personal accounts, GitHub documents a Bundled AI credits budget of $0 with Stop usage when budget limit is reached enabled to prevent extra usage charges. This is not the same as setting an organizational user-level budget to zero.
An organizational user-level budget covers all credit consumption, including included credits. At $0 it blocks credit-consuming features immediately. Organization, cost-center and enterprise spending limits instead cap metered charges and need the stop-usage setting enabled to enforce the limit.
Inline code completions and next-edit suggestions remain outside AI-credit billing on paid plans.
Before choosing Hydrogen, check your billing eligibility, model policy and extra-spend setting. Developers considering direct Google access can read RohitAI’s Gemini 4 Hydrogen API guide; that is a separate service with its own access and billing terms.