Skip to main content

Claude Fable 5 Is Back: Restored Access, Usage Terms & What Changed

Claude Fable 5 is available again as of July 1. Here is the restoration schedule, usage terms by plan, and what changed in the redeployed version.

ByUpdated

Claude Fable 5 is available again worldwide as of July 1, 2026, roughly three weeks after US government export controls (government-imposed restrictions on who can use or receive a technology) forced Anthropic to suspend access on June 12. This guide covers the restoration schedule, where and under what conditions you can use it, and what changed in the redeployed version, based on Anthropic's official announcements.

Update (November 11, 2026): Fable 5 has been followed by Fable 5.1 (September 1, 2026) and Claude Fable 5.2 (November 9, 2026), and the official model docs now list Fable 5 and Fable 5.1 as legacy. Plan terms are unchanged, and API pricing dropped to $8/$40 on Fable 5.2 (Fable 5 and 5.1 remain $10/$50). The article below describes Fable 5 — see the Claude Fable 5.2 guide for what changed.

🟢 Claude Fable 5 · model notice · Fable subscription
🟡 This article reflects Claude Fable 5. As of November 2026, the current top model is Claude Fable 5.2, and the full lineup is Claude Opus 5.5 / Claude Sonnet 5.5 / Claude Haiku 5.5 (higher tier: Claude Fable 5.2).

Fable subscription (updated November 17, 2026): Claude Fable 5.2, released November 9, 2026, is the current Fable model; Fable 5.1 and Fable 5 are now legacy. Plan terms are the same for all three — Max and Ultra plans and premium seats on Team and seat-based Enterprise plans include Fable at up to 50% of the weekly usage limit; Pro plans and standard seats use usage credits (not available on Free). API pricing is $8/M input and $40/M output tokens for Fable 5.2, and $10/$50 for Fable 5.1 and 5. Some cybersecurity and biology requests may be answered by an Opus model due to safety classifiers. See the Fable 5.2 launch news and the Fable 5.2 guide for details.

Claude Fable 5 · From Suspension to Redeployment Jun 9 Launch Jun 12 Suspended by export controls ~3 weeks offline Jun 30 Controls lifted Jul 1 Back globally Claude.ai · API Code · Cowork Deadline Plan inclusion ends Source: Anthropic official announcement (June 30, 2026)

This article reflects the situation as of July 8, 2026, based on Anthropic's official announcement (“Redeploying Claude Fable 5”, June 30) and official guidance. Terms and included benefits may change, so please check official documentation for the latest details.

What Happened — From Suspension to Restoration

Claude Fable 5 launched on June 9 as the first generally available model of the Claude 5 generation. On Friday, June 12, the US government applied export controls to Fable 5 and Mythos 5, requiring restrictions on access by foreign nationals. Because the order took effect immediately and there was no reliable way to verify user nationality in real time, Anthropic suspended both models for all users.

The trigger was a jailbreak report (a technique that makes a model bypass its safeguards) from Amazon researchers: the prompt got the model to identify several software vulnerabilities, and in one case produce code demonstrating how a vulnerability could be exploited. After reviewing the evidence with the government and partners, Anthropic officially stated that less capable models — including Claude Opus 4.8, GPT-5.5, and Kimi K2.7 — could identify the same vulnerabilities, and every model they tested could reproduce the exploit demonstration. According to press reports, Mythos 5 access was restored to a set of government-approved US organizations on June 26.

On June 30, the export controls were lifted, Anthropic officially announced the redeployment, and Fable 5 returned to users globally on Wednesday, July 1. For a detailed breakdown of the suspension and review process, see our analysis of the redeployment announcement.

Fable 5 — from suspension to returnJun 9LaunchFirst publicClaude 5 modelJun 12SuspendedUS export controls,effective at onceJul 1BackRedeployed with newsafety classifierLimited offerIn plansPro·Max·Team, up to50% of weekly limitAfter windowCreditsContinues viausage credits

Where You Can Use It Now

Per the official announcement, Fable 5 is available from July 1 in four places: Claude.ai (web and apps), the Claude Platform (API), Claude Code, and Claude Cowork. Availability through AWS, Google Cloud, and Microsoft Foundry will be re-enabled “as quickly as possible”; no exact date has been officially announced.

In Claude Code, select Claude Fable 5 with the /model command and you are ready to go.

Usage Terms and Pricing

The restoration came with a time-limited inclusion offer whose end date has been officially extended several times (extensions apply automatically). For the terms currently in effect, see the notice at the top of this article. The terms as announced:

  • Included in subscriptions (time-limited — see the notice above for the current deadline): On Pro, Max, Team, and select Enterprise plans, Fable 5 draws from your plan's usage with no additional cost — capped at up to 50% of your weekly usage limits.
  • After the inclusion window ends: You can keep using it with usage credits (a pay-as-you-go balance separate from plan limits), enabled under Settings > Usage on claude.ai.
  • API pricing: $10 per million input tokens / $50 per million output tokens.
  • Data retention: Inputs and outputs are retained for 30 days to support Anthropic's safety work.

What Changed in the Redeployed Version

This is not a simple “switch back on.” Anthropic officially stated that the redeployed model ships with a new safety classifier (a separate safety system that inspects requests and responses for risk) targeting the reported bypass technique.

  • The classifier blocks the reported technique in over 99% of cases. Blocked requests are not refused outright — they are routed to Claude Opus 4.8 instead, and the user is notified when this happens.
  • In the near term, false positives mean some routine coding and debugging requests may also fall back to Opus 4.8. Anthropic says it will refine the classifiers over the coming weeks to reduce false positives.
  • CAISI (the US Commerce Department's Center for AI Standards and Innovation) tested both the previous and new safeguards.

Anthropic is also drafting an industry framework with Amazon, Microsoft, Google, and other partners for assessing the severity of AI jailbreaks. See our safeguards and jailbreak framework analysis for details.

Getting the Most Out of It in Claude Code (Official Tips)

Anthropic's official guidance alongside the restoration lists four tips:

  • Give it the goal, not the steps: Describe the outcome and let the model plan the path. Combined with the /goal command, it keeps working until a completion condition is met.
  • Hand off long asynchronous work: It is designed to hold multi-hour and multi-day sessions without losing the thread, so you can delegate tasks you would normally break up.
  • Skip guardrail prompts: You do not need notes like “make sure to test.” It is built to review its own work; qualitative feedback works better than itemized instructions.
  • Hand it ambiguous problems: Root-causing, outage debugging, architecture calls — it more reliably identifies underlying causes instead of symptoms.

This article is for informational purposes and is based on Anthropic's official announcements. usingclaude.com is an independent information site not affiliated with Anthropic.

Was this helpful?

Keep reading