What's new in Claude Sonnet 5

<p><strong><a href="https://platform.claude.com/docs/en/about-claude/models/whats-new-sonnet-5">What&#x27;s new in Claude Sonnet 5</a></strong></p> Claude Sonnet 5 came out <a href="https://www.anthropic.com/news/claude-sonnet-5">this morning</a>. I always head straight for the "what's new" developer docs because they tend to have more actionable information than the official announcement post.</p> <p>Anthropic say of Sonnet 5 that "its performance is close to that of Opus 4.8, but at lower prices". The <a href="https://www-cdn.anthropic.com/9e6a1044980d8c4ed85669faf9c2a8342e2e9f1e/Claude%20Sonnet%205%20System%20Card.pdf">system card</a> helps explain how they were able to release the model without being blocked by the US government:</p> <blockquote> <p>Sonnet 5 is significantly less capable at cyber tasks than Mythos 5: its safeguards are thus similar to those we apply to Opus 4.7 and Opus 4.8 (models that are more capable than Sonnet 5 but much less capable than Mythos 5).</p> </blockquote> <p>Of note from the "what's new" API changes:</p> <ul> <li>Sampling parameters <code>temperature</code>, <code>top_p</code>, <code>top_k</code> are no longer supported.</li> <li>It has a 1 million token context window and 128,000 maximum output tokens.</li> <li>It features "the same set of tools and platform features as Claude Sonnet 4.6"</li> <li>Adaptive thinking is on by default, unless you specify <code>"thinking": {type: "disabled"}</code>.</li> <li>The pricing is the same as Sonnet 4.6: $3/million input, $15/million input, with an introductory discount to $2/$10 until 31st August. But...</li> <li>The model has a new tokenizer, where "The same input text produces approximately 30% more tokens than on Claude Sonnet 4.6." - effectively a 30% price increase.</li> </ul> <p>I used my <a href="https://tools.simonwillison.net/claude-token-counter">Claude Token Counter</a> tool to try out the new tokenizer. Here are my results for several larger documents:</p> <table> <thead>