Commerce Lifts Export Controls on Claude Fable 5 and Mythos 5

Anthropic announced the public release of its latest model, Claude 3.5, via a tweet on its official account. The update notes that the new version is now available to all API users and highlights several technical upgrades, including a larger context window, improved safety mitigations, and reduced inference cost. The company also mentioned that Claude 3.5 will be offered in a “Turbo” variant for customers who need higher throughput.

The announcement comes amid a broader push by AI firms to balance performance with responsible use. Claude 3.5’s expanded token limit—reported to be up to 200 k tokens—allows developers to work with longer documents and more complex conversations without truncation. Anthropic’s safety team claims the model incorporates new alignment techniques that reduce hallucinations and mitigate the generation of disallowed content, a feature that has drawn attention from both industry analysts and the HN community, which discussed the update in a thread that amassed 213 points and 78 comments.

Industry observers see the release as a strategic move to keep Anthropic competitive against larger incumbents such as OpenAI and Google. By offering a cost‑effective, high‑capacity model, the company aims to broaden its appeal to enterprises that require robust, safe AI assistants for tasks ranging from coding support to customer service. The rollout is expected to spur further experimentation and integration, with Anthropic already planning additional refinements to its safety framework in the coming months.