Anthropic ends temporary Claude Fable 5 rollout, updates subscription access
Anthropic is folding Claude Fable 5 into Max and Team Premium plans at 50% of limits from July 20
Anthropic is drawing a line under weeks of on-again, off-again access to Claude Fable 5. Starting July 20, the model becomes a standard part of Max and Team Premium subscriptions, though the fine print is worth reading closely.
Fable 5 is currently included in both Max and Premium plans at 50% of the limit, providing subscribers with a set limit rather than availability.
However, the actual limits themselves have changed on the same day: the increase that had been in effect in weekly rate limits for a while has ended on July 20, leaving the 50% amount smaller.
There is no inclusion of Fable 5 in the subscriptions for Pro and Teams plans.
They would need to continue with the usage credits, and Anthropic has granted them a one-off $100 credit as compensation for the transition period.
When that credit runs out, then the billing for the usage of Fable 5 will follow at the normal API rates, some of the most expensive of all Claude's models at present.
Anthropic claims that the constant extensions were not a method to reduce the usage but were to balance up with unpredictable demands.
That was not the only problem with Fable 5 during the month, as it had gone offline once in June due to US export control rules but had returned to service in July along with its sibling Claude Mythos 5.
-
France to impose social media ban for under-15s after parliament's approval
-
Google developing 'Frozen v2' chip to hardcode Gemini architecture with 10x efficiency
-
Microsoft adopts AMD Helios: Can it challenge Nvidia?
-
Trump administration eyes ban on Chinese AI models, including Kimi K3: Report
-
Meta's Instagram faces Tennessee addiction trial: Check details here
-
AI agents are reshaping enterprise identity security, experts say
-
'Don't buy the AI drama': Anthony Scaramucci dismisses mass layoff fears
-
China's Kimi K3 fixes 15 security bugs, challenging US AI guardrails: Report