Anthropic ends Fable 5 access questions with reduced caps and credits

Months of uncertainty around access to Anthropic’s Claude Fable 5 model have finally landed on a resolution. The company confirmed this week that Fable 5 will remain available inside its Max and Team Premium subscription tiers, but at half the normal usage caps. Lower-tier subscribers will receive a one-time $100 credit before being shifted to a pay-per-use model.

The decision ends a painful five-week stretch during which Anthropic repeatedly postponed what it called a temporary removal of Fable 5 from standard plans. The company admitted that demand proved challenging to predict and labelled the resulting access situation frustrating. In a rare moment of candour, Anthropic said it is investing in additional compute capacity to improve access going forward.

What changed

Fable 5 launched in mid-June as one of Anthropic’s flagship offerings. Within weeks, subscriber demand overwhelmed capacity, forcing the company to ration access with usage caps. The original plan was to remove Fable 5 from subscriptions entirely while capacity expanded, but that deadline was pushed back three times in five weeks.

Under the new arrangement, Max and Team Premium subscribers keep Fable 5 at half their usual caps. Users on lower plans receive a $100 credit to bridge the transition, then move to per-token pricing. Anthropic says the change provides clarity after weeks of shifting announcements.

Competitive pressure mounts

The timing is no coincidence. Moonshot AI released its Kimi K3 model last month, a near-frontier open-weight system that has quickly gained traction. At the same time, OpenAI continued expanding access to its GPT-5.6 Sol model, undercutting Anthropic on both price and availability.

OpenAI CEO Sam Altman publicly mocked Anthropic’s access struggles on X, saying clarity is welcome and implying that customers should stay with OpenAI because the company does not treat users with contempt. The jabs reflected a broader industry narrative: Anthropic’s repeated delays risk pushing enterprise and developer customers toward competitors with more predictable roadmaps.

Why it matters

For subscribers, the outcome is mixed. Fable 5 survives, which is better than the original plan to remove it entirely. Yet half-cap access and a one-time credit are unlikely to satisfy users who faced weeks of uncertainty, broken promises, and sudden workflow disruptions.

The episode exposes a growing tension in the AI industry. Model releases are outpacing compute capacity, and even well-funded labs like Anthropic are struggling to balance hype with delivery. Expect access dramas to become a recurring theme as frontier models grow more expensive to run and competition for server capacity intensifies.

Fable 5 is only six weeks old, and this access drama may well follow Anthropic from one launch to the next. The real question is whether the company can scale compute fast enough to restore confidence, or whether users will vote with their subscriptions and migrate to models with fewer usage surprises.

Related Reading

The views expressed on this site are my own and do not represent those of any current or former employer. Articles are based on publicly available information and are provided for general educational purposes.

Subscribe

Related articles

Google’s Gemini AI Autonomously Hacked Three Companies. Here’s What Happened.

Google has confirmed its Gemini AI autonomously hacked three real companies during a security test. The model guessed passwords, searched for leaked credentials, and accessed protected systems before stopping itself.

440 AI Agents Broke Into 395 Organisations in 26 Seconds. Nobody Stopped Them.

A swarm of 440 AI agents exploited two PaperCut flaws and compromised 395 organisations across 48 countries. The agents reached domain admin in 6 hours and ignored explicit instructions to stay out of 28 countries.

For $3,000 and a Few Days, Researchers Used Claude to Hack OpenAI

Security researchers used Anthropic's Claude AI to hack OpenAI's internal systems for less than $3,000 in tokens. What the HEIF Heist tells us about the new economics of cyber attacks.

The AI Hacking Crisis Is Already Here. Six New Incidents Prove It

OpenAI disclosed six new incidents where its models concealed mistakes, sought unauthorised credentials and uploaded files to the public internet. Cybersecurity experts say the real risk is powerful models meeting poor security controls.

Inside OpenAI’s Log of Misbehaving Models: Rewriting Jailbreaks and Covering Up Errors

OpenAI published six new reports of its models rewriting jailbreak instructions and concealing errors during training, alongside a faster public disclosure framework.
Philip Hall
Philip Hall
Philip Hall is a Sydney-based Cyber AI and Automation leader with more than 30 years of technology experience and a career in cyber security dating back to 2008. His work spans cyber architecture, cloud security, threat intelligence, assurance, incident support, AI-enabled defence and the security of autonomous agents.