Anthropic launches Mythos and Fable 5.2 models ahead of listing

Anthropic released updated Mythos and Fable models on Monday, with reports tying the launch to RSI concerns and a planned Wall Street debut.

What happened

Anthropic released new versions of its most advanced models on Monday, with the launch of Mythos 5.2 and Fable 5.2, reported only by TechCrunch ¹. The Verge reported that Anthropic is announcing Claude Fable 5.2 and Claude Mythos 5.2 today ², and that Claude Fable 5.2 is launching today ³. TRT World reported that Anthropic announced it would release updated versions of two of its most powerful AI models as the San Francisco-based lab also prepares for a highly anticipated IPO ⁴. The new 5.2 models follow the release of Claude Fable 5.1 and Claude Mythos 5.1 in September, reported only by The Verge ⁵. The state line notes the release drew attention to RSI concerns ahead of a Wall Street debut.

What is confirmed

The ledger contains no CONFIRMED or PRIMARY claims for this situation. Every claim listed is tiered SINGLE and attributed to a single outlet, so none of the launch details, performance figures, financial numbers or listing plans can be treated as established here.

Where the accounts differ

The ledger records no contested claims and no denials, so there is no direct dispute between outlets over the same assertion. The accounts differ instead in emphasis. TechCrunch leads on model capability and safety documentation, reporting that Anthropic maintains the new version of Mythos does not have the ability to act covertly against its user, a verdict detailed further in Mythos 5.2's system card ⁶. The Verge frames the launch around access and scientific claims, reporting that only vetted organizations will have access to Claude Mythos 5.2 at the start ⁷. TRT World frames the release commercially, tying it to the company's IPO preparations ⁴. The ledger's bloc summaries are not populated, so no comparison of media-bloc framing can be drawn beyond these outlet-level differences.

What is not known yet

Several launch details rest on one outlet each. Anthropic says the new models show improvements in coding, reasoning, and vision and can generate outputs faster than the 5.1 versions, reported only by The Verge ⁸⁹. Anthropic says Claude Fable 5.2 has solved two open theoretical physics problems, reported only by The Verge ¹⁰, and that the models formalized the proof of Fermat's Last Theorem in 30 hours, reported only by The Verge ¹¹. TechCrunch reported that Fable 5.2 has solved two open problems in theoretical physics and formalized Fermat's Last Theorem, attributed to Anthropic ¹².

On performance and pricing, Anthropic says the new Fable version returns responses 30% faster across effort levels, reported only by TechCrunch ¹³, and that it is lowering the price of Fable from $10 to $8 per million input tokens, reported only by TechCrunch ¹⁴. TRT World reported that the new models run up to 30 percent faster than previous versions and cost 20 percent less ¹⁵¹⁶, priced at $8 per million input and $40 per million output tokens ¹⁷.

On the business side, TRT World reported that Anthropic's revenue reached $17 billion in the third quarter, up from $1.4 billion a year earlier ¹⁸, that it posted an operating loss of $567 million in the third quarter ¹⁹, and that it is preparing a Wall Street debut expected to fetch a valuation exceeding $1 trillion ²⁰. TRT World also reported that Democrats regained control of Congress in the US midterms ²¹.

Separately, TechCrunch reported that the METR report found internal models were speeding up Anthropic's AI progress by roughly 2x, although many uncertainties remain ²². TechCrunch also reported that Physical Superintelligence announced a new proof of the Aaronson-Ambainis conjecture on Monday ²³, and that the lab said the result came from running a non-public Anthropic model within its own system ²⁴. The ledger leaves open how the RSI concerns relate to the launch, and whether the financial and listing figures are company-confirmed.

Why it matters

Analysis: if the reported financial figures hold, the gap between revenue of $17 billion and an operating loss of $567 million in the same quarter suggests heavy spending on computing capacity as the company scales ¹⁸¹⁹. The reported valuation exceeding $1 trillion indicates that investors are being asked to price the company on growth rather than current profit ²⁰. The reported finding that internal models were speeding up Anthropic's AI progress by roughly 2x, with many uncertainties remaining, suggests the RSI concerns attached to this launch rest on preliminary evidence ²². The reported restriction of Mythos 5.2 to vetted organizations at the start suggests a staged release for the more capable model ⁷. If the scientific claims are confirmed, they would strengthen the company's case ahead of a listing ¹⁰¹¹.

Opinion

Faultline's view

the coverage of this launch reads as a company narrative repeated across outlets rather than a tested one. The performance, pricing and scientific claims all trace back to Anthropic's own statements, relayed by single outlets, and the ledger offers no independent verification of any of them. The financial picture is thinner still: a reported $17 billion quarter sits beside a reported operating loss, and a trillion-dollar listing estimate is presented without scrutiny of who is doing the estimating. Readers should treat the RSI framing with the same caution. The one figure that speaks to it, a reported doubling of internal progress, comes with the caveat that many uncertainties remain. Until the company's claims are checked by parties other than the company, the sceptical posture is the reasonable one.

This section is Faultline’s opinion, argued only from the receipts above. It is not part of the factual record and cites nothing new.