--- title: 7 ways I make Claude Code use fewer tokens without... description: Practical tips for reducing Claude Code's token consumption without sacrificing output quality. Techniques include starting fresh sessions with /clear instead... canonical: https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: 7 ways I make Claude Code use fewer tokens without hurting the results | daily.dev og:description: Practical tips for reducing Claude Code's token consumption without sacrificing output quality. Techniques include starting fresh sessions with /clear instead... og:url: https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r og:image: https://api.daily.dev/og/posts/S1mmfwQ7r.png og:image:alt: 7 ways I make Claude Code use fewer tokens without hurting the results og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # 7 ways I make Claude Code use fewer tokens without hurting the results **[XDA Developers](https://daily.dev/sources/xda-developers)** · 7 min read · 1 upvotes · 0 comments ## Summary Practical tips for reducing Claude Code's token consumption without sacrificing output quality. Techniques include starting fresh sessions with /clear instead of dragging context forward, using /context to identify what's consuming the context window (system tools often eat more tokens than actual messages), running /compact to condense long conversations, lowering the effort level from High to Medium (roughly a 45% token drop with no quality loss), matching the model (Haiku, Sonnet, Opus, Fable) to task complexity, using @mentions to point Claude directly to relevant files instead of letting it search, and interrupting the agent immediately when it goes off track rather than letting it burn tokens on a wrong approach. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How do I reduce token usage in Claude Code without hurting the quality of results? Use /clear between unrelated tasks to wipe stale context, run /context to see what's consuming tokens, use /compact to condense long sessions, lower the effort level from High to Medium, match the model (Haiku, Sonnet, Opus, Fable) to task complexity, use @mentions to point directly to relevant files, and interrupt Claude the moment it goes off track. _Developers tuning agent workflows can compare token-saving techniques for Claude Code on daily.dev._ ### Does lowering the effort level in Claude Code reduce token usage significantly? Yes, switching the effort level from High to Medium in Claude Code produced roughly a 45% drop in token usage on identical tasks, with no meaningful difference in result quality. Anthropic's newer models use Adaptive Reasoning, letting Claude decide how deeply to reason per step, but the chosen effort level still shapes how readily it reasons. _Anyone managing AI coding costs can track effort-level tradeoffs like this on daily.dev._ ### What is eating up my context window tokens in Claude Code even before I send a message? System tools, built-in agents, skills, and connected MCP servers can consume tens of thousands of tokens before any user message is counted. In one breakdown, total usage was about 41,000 tokens, with only 2,600 from actual messages and over 20,000 taken up by system tools alone, including unused connected MCP servers. _Developers debugging bloated context windows can dig into fixes like this on daily.dev._ ## Similar posts on daily.dev - [I stopped giving Claude Code my entire project, and my Claude usage lasted 4x longer](https://daily.dev/posts/i-stopped-giving-claude-code-my-entire-project-and-my-claude-usage-lasted-4x-longer-0o1ubxjob) · XDA Developers · 0 upvotes · 0 comments - [Maximize AI Efficiency: Reduce Token Use in Claude Pro](https://daily.dev/posts/maximize-ai-efficiency-reduce-token-use-in-claude-pro-vwm1wdmka) · Sergio Lema · 1 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#mcp](https://daily.dev/tags/mcp), [#anthropic](https://daily.dev/tags/anthropic), [#claude-code](https://daily.dev/tags/claude-code), [#prompt-engineering](https://daily.dev/tags/prompt-engineering) [View this post on daily.dev](https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"7 ways I make Claude Code use fewer tokens without hurting the results","url":"https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r"},"datePublished":"2026-08-20T18:03:07.684Z","dateModified":"2026-09-14T07:07:26.886Z","description":"Practical tips for reducing Claude Code's token consumption without sacrificing output quality. Techniques include starting fresh sessions with /clear instead...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/271d5db869f59988c24cbd777302109b?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/271d5db869f59988c24cbd777302109b?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"XDA Developers","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"XDA Developers","logo":"https://media.daily.dev/image/upload/s--kCg6yyAP--/f_auto,q_auto/v1774964407/logos/xda-developers?_a=BAMAMiWQ0","url":"https://daily.dev/sources/xda-developers"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,mcp,anthropic,claude-code,prompt-engineering","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"XDA Developers","item":"https://daily.dev/sources/xda-developers"},{"@type":"ListItem","position":3,"name":"7 ways I make Claude Code use fewer tokens without hurting the results"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/7-ways-i-make-claude-code-use-fewer-tokens-without-hurting-the-results-s1mmfwq7r#faq","mainEntity":[{"@type":"Question","name":"How do I reduce token usage in Claude Code without hurting the quality of results?","acceptedAnswer":{"@type":"Answer","text":"Use /clear between unrelated tasks to wipe stale context, run /context to see what's consuming tokens, use /compact to condense long sessions, lower the effort level from High to Medium, match the model (Haiku, Sonnet, Opus, Fable) to task complexity, use @mentions to point directly to relevant files, and interrupt Claude the moment it goes off track. Developers tuning agent workflows can compare token-saving techniques for Claude Code on daily.dev."}},{"@type":"Question","name":"Does lowering the effort level in Claude Code reduce token usage significantly?","acceptedAnswer":{"@type":"Answer","text":"Yes, switching the effort level from High to Medium in Claude Code produced roughly a 45% drop in token usage on identical tasks, with no meaningful difference in result quality. Anthropic's newer models use Adaptive Reasoning, letting Claude decide how deeply to reason per step, but the chosen effort level still shapes how readily it reasons. Anyone managing AI coding costs can track effort-level tradeoffs like this on daily.dev."}},{"@type":"Question","name":"What is eating up my context window tokens in Claude Code even before I send a message?","acceptedAnswer":{"@type":"Answer","text":"System tools, built-in agents, skills, and connected MCP servers can consume tens of thousands of tokens before any user message is counted. In one breakdown, total usage was about 41,000 tokens, with only 2,600 from actual messages and over 20,000 taken up by system tools alone, including unused connected MCP servers. Developers debugging bloated context windows can dig into fixes like this on daily.dev."}}]} ```