Claude Fable 5.2 in GitHub Copilot: Admin Opt-In, Pricing

Nothing new will appear in your Copilot model picker until someone with admin rights says yes. On Nov. 9, 2026, GitHub announced that Claude Fable 5.2 is generally available in GitHub Copilot, and for Business and Enterprise plans the model ships switched off.

GitHub's post on X, Nov. 9, 2026, announcing Claude Fable 5.2 in Copilot (@github)

View on X

GitHub's post says its early testing found Fable 5.2 "solved more tasks than Fable 5.1 while using fewer steps and tool calls" and "completed difficult tasks that earlier models missed." It recommends the model for complex, multistep coding jobs across large codebases, in the Copilot app, the CLI and VS Code. GitHub gives no task counts or scores, so treat that as a vendor claim; Anthropic's own benchmark table is in our breakdown of the Fable 5.2 launch.

Who gets it, and where

The changelog names Copilot Pro+, Max, Business and Enterprise. It doesn't name the Free or Pro plans. You can pick the model in Visual Studio Code, Visual Studio, Copilot CLI, the Copilot coding agent, the Copilot app, github.com, GitHub Mobile, JetBrains IDEs, Xcode and Eclipse. Rollout is gradual, so some accounts won't see it yet.

The switch that accepts data retention

Business and Enterprise administrators must enable a "Claude Fable 5.2" policy in Copilot settings, and it is off by default. GitHub spells out the stakes:

Enabling the Claude Fable 5.2 policy constitutes acknowledgment of this requirement. Leaving it off keeps Claude Fable 5.2 unavailable to your organization.

The requirement is that Anthropic keeps prompts and outputs so its safety classifiers can run. GitHub says that data isn't used to train Anthropic's models, and that other Claude models, apart from Fable 5, 5.1 and 5.2, still run with zero data retention. Anthropic's help-center policy puts the retention at 30 days, says it has applied since June 9, 2026, and says flagged content or legal holds can extend it.

What it costs per token

GitHub bills the model "at provider list pricing under usage-based billing." Its pricing page lists Fable 5.2 at $8.00 per million input tokens, $0.20 for cached input and $40.00 for output. Fable 5.1 is listed at $10.00, $0.25 and $50.00, and Opus 5.5 is listed at half of Fable 5.2 on input and output.

By our math, Fable 5.2 is 20% cheaper than 5.1 and costs exactly twice what Opus 5.5 does ($4 in, $20 out) (calculations are ours, checked twice). Cached input is the same $0.20 for both newer models, which matters for agents that resend large contexts.

The year-end deadline for zero retention

Eligible enterprises can run Fable 5.2, 5.1 and 5 with zero data retention "through the end of the calendar year" under a time-bound exemption while Anthropic rolls out Enterprise Frontier Safeguards. After that, continued use will require those safeguards, which GitHub says will bring automated safety monitoring and options such as customer-controlled encryption keys. Only GitHub account teams can say who qualifies, and approval doesn't switch the model on: an admin still has to enable the policy.

Our practical read: try Fable 5.2 on the hardest tickets and keep Opus 5.5 for routine work, since the price gap is 2x. We couldn't find what Enterprise Frontier Safeguards will cost or require beyond that description. If your legal team blocks retained prompts, ask GitHub for your eligibility before Dec. 31, because that is the date the current exemption stops covering you.

Source: GitHub Changelog, Claude Fable 5.2 in GitHub Copilot

github copilot claude fable 5 2 anthropic data retention ai coding

AI

About the author

AI Briefing Daily Editorial. The editorial desk of AI Briefing Daily, an independent publication covering artificial intelligence. Every story is researched against primary sources, checked by hand and reviewed by a person before publication.