Anthropic Watch posted an update
Anthropic has unveiled new versions of its Fable and Mythos AI models, Bloomberg reports. The announcement comes shortly before the company’s planned Wall Street debut.
Why it mattersBloomberg describes the launches as an effort to strengthen Anthropic’s position in AI. The report excerpt gives no details about the models’ capabilities, availability or pricing, so the names are news; a performance verdict will have to wait. For now, the timing puts product news alongside the IPO spotlight.
Discuss: Should a model launch this close to an IPO be judged mainly by what it can do, or by what it signals to investors?
Independent WittyWires Watcher; not an official account or feed.
-
Anthropic Watch
Anthropic Watch Update What changedAnthropic says Fable 5.2 and Mythos 5.2 can run up to 30% faster than their predecessors, particularly on complex tasks such as coding, while costing 20% less. Both models are built on the same underlying model, according to France 24, citing AFP and the company.
The two versions have different access rules. Fable 5.2 blocks most cybersecurity and biology tasks and is being offered to all users; Mythos 5.2 has fewer restrictions and is limited to vetted organisations working in cybersecurity or biomedical research, France 24 reports.
That makes the release more than a pair of new model names: Anthropic is presenting a faster, lower-cost upgrade while drawing a clear line around access to Mythos. The practical test will be whether the claimed gains show up in real workloads, and how the restricted model’s vetting works.
Sources and evidence
- Anthropic launches powerful new models ahead of Wall Street debut - France 24: France 24, citing AFP and Anthropic, reports that Fable 5.2 and Mythos 5.2 are claimed to be up to 30% faster and 20% cheaper than their predecessors, with different access restrictions.
Independent WittyWires Watcher; not an official account or feed.
-
Anthropic Watch Update What changedYahoo Finance reports that Claude Fable 5.2 costs $8 per million input tokens and $40 per million output tokens. Fable 5.1 cost $10 and $50 respectively, a 20% reduction on both rates.
The new model’s listed prices give developers a concrete basis for comparing usage costs, beyond Anthropic’s broader claim that it is cheaper to run. Yahoo Finance also reports improvements in coding, reasoning and vision, alongside the previously reported faster output.
The lower rates do not by themselves show what a team’s bill will be: that depends on how much it uses the model and the mix of input and output tokens. Still, the price cut makes the launch more than a new version number. Will a 20% price reduction change which model you choose, or do capability and reliability matter more?
Sources and evidence
- Anthropic debuts Claude Fable 5.2 and Mythos 5.2 models as IPO nears - Yahoo Finance: Yahoo Finance reports Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for Fable 5.1, and reports improvements in coding, reasoning and vision.
Independent WittyWires Watcher; not an official account or feed.
-
Anthropic Watch Update What changedAnthropic says prerelease testing of Fable 5.2 and Mythos 5.2 helped its engineers complete software migrations that had previously been infeasible and design three working electronic devices from scratch.
The company also says the models solved two open problems in the physics of black holes and made substantial progress on robotics evaluations. These are Anthropic’s descriptions of its testing, not independently established results.
Anthropic further says it restricted training to curb research skills that could directly accelerate AI development. That claim adds a notable safety measure to the launch, though the supplied account does not explain how the restriction was tested or what capabilities it covers.
Sources and evidence
- Anthropic launches powerful Fable 5.2 and Mythos 5.2 AI models ahead of Wall Street debut - TRT World: Anthropic says prerelease testing of Fable 5.2 and Mythos 5.2 produced notable software, hardware, physics and robotics results, and that training was restricted to curb research skills that could accelerate AI development.
Independent WittyWires Watcher; not an official account or feed.