GitHub Copilot: Claude Fable 5 generally available for all premium and business plans
Anthropic's Claude Fable 5, the first model of the newly established Mythos class designed for long-horizon autonomous tasks, is generally available in GitHub Copilot as of June 9, 2026, for Pro+, Max, Business, and Enterprise plans — with specific notes on data retention.
This article was generated using artificial intelligence from primary sources.
GitHub Copilot announced the general availability (GA) of Claude Fable 5 on June 9, 2026, for all premium and business subscription plans. This is the first model of Anthropic’s newly established Mythos class — a line of models designed for long-horizon, autonomous tasks in coding and knowledge work.
What is the Mythos class and what is Fable 5 for?
Anthropic introduced a new model category for Fable 5 called the Mythos class. While previous Claude lines — Opus, Sonnet, and Haiku — were optimized for different speed-to-quality trade-offs, Mythos is designed from the ground up for long-horizon, autonomous tasks where an agent must independently plan, research, write, and fix code across multiple steps, without constant human supervision.
That architecture is precisely why GitHub chose Fable 5 as the new flagship model in the Copilot ecosystem: the model handles long-horizon tasks — such as refactoring an entire module, writing tests for a legacy codebase, or researching and implementing a new API integration layer — better than any previous Claude model.
Performance: fewer tool calls, lower token consumption
GitHub conducted internal measurements comparing Fable 5 with previous Opus-tier models. Results show that Fable 5 completed equivalent tasks with fewer tool calls and lower overall token consumption.
For users, this has a concrete practical consequence: complex, multi-step tasks can be completed faster and at a lower cost per request. Billing is usage-based per Anthropic’s pricing — no flat surcharge on top of the subscription, only what is generated.
Availability by plan and platform
Claude Fable 5 is available starting today for users on the following plans:
- Copilot Pro+
- Copilot Max
- Copilot Business
- Copilot Enterprise
The model works across all major development environments that Copilot supports: Visual Studio Code, Visual Studio, JetBrains IDEs, Xcode, and Eclipse, as well as in Copilot CLI and the GitHub Copilot cloud agent. On the web it is integrated directly into github.com, and mobile users can access it through the GitHub Mobile app for iOS and Android. Copilot Free plan users do not have access to this model.
Data retention — a special requirement
The most important difference between Fable 5 and the other Claude models in Copilot concerns data handling. All existing Claude models available in Copilot — Opus 4.8, Sonnet 4.6, and Haiku 4.5 — operate under Zero Data Retention: queries and responses are not stored at GitHub or Anthropic after the session ends.
Fable 5 is an exception. Due to the architecture of Anthropic’s safety classifiers, prompts and outputs are retained for up to 30 days. Anthropic explicitly states that this data is not used for model training, but users and organizations should take this requirement into account when handling sensitive business, user, or regulated information.
Administrative settings for business users
For organizational plans — Enterprise and Business — Fable 5 is not automatically enabled. Administrators must explicitly enable the appropriate Copilot policy in the management dashboard before users in the organization can select this model.
GitHub made this choice deliberately: since the 30-day data retention requirement is incompatible with some regulatory frameworks or internal policies, the activation decision is left to administrators. Individual plan users — Pro+ and Max — can immediately select Fable 5 as the active model without any additional steps.
Claude Fable 5 marks Anthropic’s entry into the segment of models that operate autonomously over a longer time horizon — and GitHub is the first major partner to integrate it at the production level. For development teams using Copilot Enterprise or Business, the key step is evaluating the data retention requirement and, where appropriate, enabling the model through the administrator settings.
Frequently Asked Questions
- What is Anthropic's Mythos class and how does Fable 5 differ from Opus models?
- Mythos is a new line of Claude models designed for long-horizon, autonomous coding and knowledge work tasks. Unlike Opus-tier models, Fable 5 completes equivalent tasks with fewer tool calls and lower token consumption according to GitHub's internal measurements.
- Is there a difference in data retention between Fable 5 and other Claude models in Copilot?
- Yes — Opus 4.8, Sonnet 4.6, and Haiku 4.5 operate under zero data retention, while Fable 5 requires up to 30 days of prompt and response storage due to Anthropic's safety classifiers. Data is not used for model training.
- On which platforms is Claude Fable 5 available in GitHub Copilot?
- The model works within Visual Studio Code, Visual Studio, JetBrains IDEs, Xcode, Eclipse, Copilot CLI, the GitHub Copilot cloud agent, github.com, and the GitHub Mobile app for iOS and Android.
📬 AI news in your inbox
A daily digest built your way — pick topics, sources and cadence. One-click unsubscribe.