GitHub Copilot adds Gemini 4 Argon on list-price billing for paid plans
Gemini 4 Argon, Google's newest frontier model, is now generally available inside GitHub Copilot on the Pro+, Max, Business and Enterprise plans, billed at provider list pricing under usage-based billing.

Image: GitHub
Why it mattersA Copilot team that was already paying for one of the four higher plans can switch to Argon in the picker today, but the token cost falls on usage-based billing rather than the plan's included quota.
A Copilot subscriber who opens the model picker in Visual Studio Code this week sees a new entry at the top. GitHub announced on 13 October that Gemini 4 Argon, Google's newest frontier model, is now generally available inside GitHub Copilot across the Pro+, Max, Business and Enterprise plans.
The model is billed at provider list pricing under Copilot's usage-based billing, GitHub says, with a link to the plan's own Models and pricing page for the exact numbers. Argon is Google's own naming for the model Google DeepMind launched on 30 September and moved to general availability on 13 October at two dollars per million input tokens and ten dollars per million output tokens on the Gemini API.
Where the picker shows it
GitHub lists seven surfaces where the model appears: Visual Studio Code, Visual Studio, the Copilot CLI, the GitHub Copilot coding agent, the GitHub Copilot app, JetBrains IDEs, and Xcode and Eclipse. Rollout is gradual, the changelog says, so a developer who does not yet see Argon in the picker is being asked to check again later rather than contact support.
Plans and admin controls
Argon sits with the frontier-class models Copilot already exposes to its paid tiers, so Copilot Free and Copilot Pro are not on the list. For a team on Copilot Business or Enterprise, admins can manage access through the model policy in Copilot settings, and new models are enabled by default under default model enablement unless an admin has turned that global off or has explicitly disabled Argon. The changelog says this in two sentences and does not change the policy in any other way.
Google's own description of what Argon is for lands the same way GitHub reports it: complex coding, agentic and reasoning tasks. GitHub adds its own early-testing claim, that Argon solved more tasks than previous Gemini models on every coding benchmark it ran, with its largest gains in Copilot CLI, and that it tests edge cases before it finishes a task. Those are GitHub's own numbers against its own earlier Gemini versions. No independent benchmark is cited in the entry.
For a Copilot admin on Business or Enterprise, the decision for the next few days is whether to leave Argon enabled or turn it off, because the model is on the usage-based meter rather than the flat seat price. For an individual on Pro+ or Max, the model is available, the picker is where it sits, and the only cost signal in the changelog is a link to the pricing page.
Source
GitHub Changelog, Gemini 4 Argon in GitHub Copilot. Model and pricing figures from Google, Introducing Gemini 4 Argon.
This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.
Get AI News in your inbox
New developer tools, model and agent releases, and how teams are actually using them to release software. Short, and only when there is something worth reading.


