The Claude Fable 5.2 model is available in GitHub Copilot
The Claude Fable 5.2 model from Anthropic is now generally available in GitHub Copilot for the Pro+, Max, Business, and Enterprise plans. According to the company, it solves more tasks with fewer steps than Fable 5.1, but by default it requires data storage for security checks.
GitHub has announced general availability of the Claude Fable 5.2 model from Anthropic in GitHub Copilot, designed for complex multi-step programming tasks in large codebases. The model is available in the model picker for users on Copilot Pro+, Max, Business, and Enterprise plans, with rollout happening gradually. For Business and Enterprise plans, an administrator must first enable the model in Copilot settings, where it is disabled by default. Pricing follows the provider's price list as part of pay-as-you-go billing.
According to Anthropic, the Claude Fable 5.2 model solved more tasks than the previous Claude Fable 5.1 in early testing, doing so with fewer steps and tool calls, and also handled difficult tasks on which older models failed.
The Claude Fable 5.2 model by default requires data storage (retention) so that Anthropic's security classifiers detecting malicious use can function; according to the company, stored data is not used to train models. Zero data retention (ZDR) is available only for selected enterprise customers, until the end of the calendar year, based on a time-limited exemption, before Anthropic implements Enterprise Frontier Safeguards (EFS). After this deadline, continued use of the model with zero retention will require EFS. Enabling the relevant policy in Copilot administration means agreeing to the data retention condition; leaving the policy disabled will block the model for the organization. Customers must request an eligibility assessment for zero retention from their GitHub account team; GitHub support does not decide on this.
Why it matters
Developers now have an additional model option in Copilot that, according to Anthropic, handles more complex programming tasks with fewer steps. For companies, especially in regulated industries, the key decision concerns data retention: the default setting stores prompts and outputs for security checks, and anyone who needs zero retention must verify eligibility and prepare to transition to Enterprise Frontier Safeguards by the end of the year.
Two audiences, two different impacts
What this means
For individuals
Developers using GitHub Copilot can now select the Claude Fable 5.2 model for complex multi-step programming tasks directly in the model picker.
For a business
Copilot Business and Enterprise administrators must explicitly enable the model and consider the impact of the default data retention for security classifiers; companies requiring zero retention have only a transitional period until the end of the year before Enterprise Frontier Safeguards will need to be implemented.
DevelopmentCheck the original
Event sources
clearly official source · 1 publisher, 0 independent. We count feeds from the same owner only once.