<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Claude-Opus-4-7 on AI Engineer Guide</title><link>https://aiengineerguide.com/tags/claude-opus-4-7/</link><description>Recent content in Claude-Opus-4-7 on AI Engineer Guide</description><generator>Hugo</generator><language>en-US</language><copyright>Copyright © 2024, Nesin Technologies LLP.</copyright><lastBuildDate>Fri, 17 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://aiengineerguide.com/tags/claude-opus-4-7/feed.xml" rel="self" type="application/rss+xml"/><item><title>Anthropic Claude Opus 4.7 uses 33% more tokens than Opus 4.6</title><link>https://aiengineerguide.com/til/anthropic-claude-opus-4-7-uses-more-tokens/</link><pubDate>Fri, 17 Apr 2026 00:00:00 +0000</pubDate><guid>https://aiengineerguide.com/til/anthropic-claude-opus-4-7-uses-more-tokens/</guid><description>&lt;p>Even though the cost of Claude Opus 4.6 &amp;amp; Claude Opus 4.7 is same ($5/M input tokens and $25/M output tokens).&lt;/p>
&lt;p>However, Claude Opus 4.7 uses &lt;strong>different tokenizer&lt;/strong> which might use up to 33% more token usage for the same input if you compare it with Opus 4.6 or other models.&lt;/p>
&lt;p>They&amp;rsquo;ve explictly called that out in their website.&lt;/p>
&lt;blockquote>
&lt;p>Opus 4.7 uses a new tokenizer compared to previous models, contributing to its improved performance on a wide range of tasks. This new tokenizer may use up to 35% more tokens for the same fixed text.&lt;/p></description></item></channel></rss>