Claude Sonnet 5 is the default model for Free and Pro users and is available on Max, Team, and Enterprise plans, in Claude Code, and through the API. Its performance is close to Opus 4.8 but at lower prices, and Anthropic is offering an introductory rate of $2/$10 through August 31 that will revert to $3/$15 afterward; the model also uses an updated tokenizer that can map the same input to roughly 1.0–1.35× more tokens depending on content type.
Claude Sonnet 5 posts a SWE-bench Pro score of 63.2%, compared with Sonnet 4.6 at 58.1%. On GDPval-AA v2, Sonnet 5 scores 1,618 while Opus 4.8 scores 1,616. On Humanity’s Last Exam, Sonnet 5 records 57.4% versus Opus 4.8 at 57.9%. These reported figures show Sonnet 5 improving over Sonnet 4.6 on SWE-bench Pro and closely matching Opus 4.8 on GDPval-AA v2 and Humanity’s Last Exam, based on the published scores.
Claude Sonnet 5 features an updated tokenizer which enhances how the model processes text. This change allows the same input to map to approximately 1.0 to 1.35 times more tokens, depending on the type of content. This improvement targets better text processing capabilities, contributing to its competitive performance.
Pricing for using this model begins with an introductory rate of $2/$10 available through August 31, after which the cost will rise to $3/$15. This pricing structure reflects the tradeoff aiming to balance cost efficiency with enhanced technological features.
An export control directive suspended access to Fable 5 and Mythos 5 for foreign nationals on June 12. The suspension applies specifically to those two models under that directive. Access for foreign nationals has been restricted since that date.
Sonnet 5 was never trained on cybersecurity tasks and scored 0% on developing a working Firefox exploit. Sonnet 5 ships with lighter safeguards than Fable’s lockdown. The model’s training scope and safeguard level differ from the suspended models.
These paragraphs summarize the reported access restrictions and model limitations. They do not add operational details beyond the published items. No further specifics are included here.
Claude Sonnet 5 is deployed as the default model for Free and Pro users and is available on Max, Team, and Enterprise plans, in Claude Code, and via the API. Its reported performance closely matches Opus 4.8 while being offered at lower prices, including an introductory rate of $2/$10 through August 31 that reverts to $3/$15 afterward. Related models Fable 5 and Mythos 5 have been suspended for foreign nationals under an export control directive since June 12.
