Using Claude Opus 5.5 as your daily driver

Summary
This video presents an overview and practical demonstration of Claude Opus 5.5 inside Claude Code, hosted by developer advocate Lydia Hallie. She highlights key performance, conciseness, and cost improvements over Claude Opus 5 and demonstrates how to optimize workflows using effort levels, subagent model configuration, and prompt auditing.

What is shown

• Side-by-side performance comparison [00:23]: A simultaneous benchmark run of Opus 5 (left) versus Opus 5.5 (right) on the same bug fix prompt ("Fix #418: refunds on orders that used discount codes come out a few cents off..."). Opus 5.5 finishes in under a minute with a concise summary and clear follow-up, while Opus 5 takes longer, makes more tool calls, and produces verbose output.
• Token and usage limit impact [01:10]: Inspection of Claude Code session limits, showing Opus 5.5 consumed ~4% of the 5-hour quota (31.4k context) compared to 6% (44.0k context) for Opus 5.
• Effort level configuration [01:27]: Demonstration of the "Effort" slider (Medium vs. High). A field rename prompt ("customerRef to accountRef") partially succeeds on Medium by only editing the handler [01:44], but on High effort [02:16], Opus 5.5 traces full dependencies across serialization files, API schemas, and test suites.
• Subagent model routing [02:34]: Setting read-only repository exploration subagents to run on Claude Sonnet instead of Opus via .claude/agents/explore.md frontmatter or the CLAUDE_CODE_SUBAGENT_MODEL variable in settings.json.
• Prompt optimization command [03:05]: Running /claude-api prompt-audit to inspect and streamline CLAUDE.md guidelines and custom skills for Opus 5.5.

Claims & numbers

• The presenter claims Claude Opus 5.5 is 20% cheaper per token than Opus 5 ($4 input, $20 output per million tokens).
• The presenter states usage limits go 25% further on Pro, Max, and Team subscriptions.
• The presenter claims tasks are approximately 40% cheaper overall due to needing fewer tokens to reach a result.
• In the side-by-side coding task demonstrated, the presenter states Opus 5.5 completed in under a minute and was 30% faster than Opus 5.
• On the Max tier 5-hour quota, Opus 5.5 used 4% of the limit versus 6% for Opus 5 on the same refund bug fix.

Notable quotes

• "5.5 is done in under a minute, and its whole response fits right here on the screen." [00:40]
• "Effort is basically how much thinking it puts into a turn before it acts." [01:30]
• "A subagent that's only exploring the codebase doesn't need Opus-level reasoning." [02:38]

Assessment
This is an official demonstration walkthrough from Anthropic highlighting Claude Opus 5.5 features in Claude Code. The presented coding tasks and side-by-side terminal sessions are shown in real software environments, though runtime test sequences are sped up for video pacing.

Described by gemini-3.8-flash on 2026-09-29 from the video's audio and frames.