Just today, Fable 5.1 has reportedly entered beta testing.
A source claims that some Claude users have already started using Fable 5.1 for free.
The first to release this news was the industry's recognized reliable leaker leo@synthwavedd.
He revealed that the successor to Fable 5 (most likely named Fable 5.1) has entered a small-scale beta test.
If you have selected Fable 5 while using the Claude Web client, or if you are using Claude Code, then you have likely already been quietly taken over by Fable 5.1.
Previously, last weekend, Anthropic released a report stating that they had internally developed a Model 2 stronger than Mythos, but it would not be released.
Fable 5.1, Imminent Full Rollout
After this news broke, multiple sources speculated it is highly likely true.
That's exactly how Fable 5 was initially released.
In early June, on the day Fable 5 was fully launched, the situation was similar to now: first testing quietly with a small group of users (about 5%), then rolling out to everyone that same day.
The purpose of beta testing is to use a small portion of real traffic as a touchstone; if it holds steady, then officially launch.
Today follows a similar pattern, indicating that the full rollout of Fable 5.1 is imminent!
The good news is: Fable 5.1 likely won't see a price increase.
It is very likely to continue the current pricing of "$10 per million input tokens, $50 for output" and continue the "more for the same price" policy.
This is really cost-effective for developers.
In short, Fable 5.1 will most likely be released before the end of August.
What's even more exciting is that beta testing and full release happening on the same day has occurred before.
The Strongest Model, Put on a Leash
Why is everyone so eager for Fable 5.1?
That day, Anthropic dropped a shocking double release: Fable 5 and Mythos 5 debuted together.
However, the complete, strongest version of Mythos 5 was only made available to a very few national security agencies by the official party; the public version Fable 5 accessible to ordinary users is the version with a "safety classifier" added.
As a result, just three days after launch, Fable 5 was urgently halted due to its excessively powerful performance, not regaining access until July 1st.
Moreover, when invoking Fable 5, if the safety classifier is triggered, the conversation is taken over by Opus 5.
Furthermore, on the official page, safety measures for Fable 5 have been continuously tightened, with the most recent update in early August.
To describe Fable 5 in one sentence—it's powerful in performance, but restricted everywhere.
Therefore, there is currently high demand from developers worldwide for Fable 5.1.
Developer @babayagatwt expressed having the following three wishes.
1. Want Fable 5.1
2. Want to remove the damned 5-hour usage limit
3. Want Fable to cover 100% of usage
Some also ask: Could Fable 5.1 just be an improved version of a restricted Mythos?
To this, leaker Pankaj Kumar replied bluntly: "Honestly, they won't release the full version of Mythos."
Moreover, some have noticed that Fable 5 has significantly become dumber over the past few weeks. Could this be precisely because Fable 5.1 is coming?
Claude Experiences Global Major Outage!
What's even more interesting is that on the same day the new model was leaked, Claude crashed globally again.
In the early hours of this morning, Anthropic's models suddenly experienced a global major outage.
The disaster struck right on time. Mythos 5, Fable 5, Opus 5, Sonnet 5, Haiku 4.5—none were spared. claude.ai, API clients, Claude Code collectively experienced disruptions for nearly 3 hours!
This is already the 166th service disruption for Claude this year.
In just the last five days, it has crashed a full 7 times: service disruption on the 16th, performance degradation on the 17th, complete outage on the 18th... almost crashing once a day.
This frequent crashing might also be related to the leak of Fable 5.1.
Why are Anthropic's servers so prone to crashing? Most likely due to insufficient computing power.
At the same time, Claude Code's 50% quota increase per week has been extended to the end of this month, acknowledging tight computing power in the coming weeks.
Just the day before the outage, this news shook Wall Street: Anthropic's annualized revenue has soared to $65 billion! A 7-fold surge from the end of last year, directly leaving old rival OpenAI's $40 billion position behind, and aiming for a trillion-dollar scale by year-end!
To handle this avalanche of fortune, Anthropic's top brass has gone completely all-out.
They have locked in approximately $71 billion in chip lease debt through an SPV structure.
Simultaneously, they just secured a $5 billion strategic investment from AMD.
In addition, they have signed multiple record-breaking large-scale data center contracts!
The ultimate goal of all this is to rush towards the nearly trillion-dollar valuation IPO window in September to October.
On one side, sprinting towards a trillion-dollar valuation; on the other, models frequently experiencing global outages. This is the current situation for Anthropic.
On the Eve of the August ASI Showdown, Cards are Revealed
Currently, Anthropic's new model is already on the verge of release, ready to go.
Facing OpenAI's step-by-step pressure, Anthropic has employed a "Tianji horse racing" tactic: hold back, then strike precisely.
This timing is also very delicate.
For Anthropic, Fable 5.1 is not a simple "version number + 0.1."
It must prove the company can digest tens of billions in computing power investment and is a trump card that must be played on the eve of the IPO.
This time, if Fable 5.1 doesn't show significant performance improvements, it might put Anthropic under even greater pressure.
There are only ten days left in August. Let's wait for Anthropic to reveal their hand.
References:
https://x.com/synthwavedd/status/2089719312340693316
https://x.com/kimmonismus/status/2089722853461483870
This article is from the WeChat public account "ASI Apocalypse", author: ASI Revelation
