--- title: Why Claude Code Time Estimates Are Poor | daily.dev description: Claude Code and similar coding agents tend to give inflated time estimates for implementation tasks because they're trained on human-written data reflecting... canonical: https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why Claude Code Time Estimates Are Poor | daily.dev og:description: Claude Code and similar coding agents tend to give inflated time estimates for implementation tasks because they're trained on human-written data reflecting... og:url: https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x og:image: https://api.daily.dev/og/posts/nBDSLhx0x.png og:image:alt: Why Claude Code Time Estimates Are Poor 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. # Why Claude Code Time Estimates Are Poor **[Towards Data Science](https://daily.dev/sources/tds)** · 8 min read · 2 upvotes · 0 comments ## Summary Claude Code and similar coding agents tend to give inflated time estimates for implementation tasks because they're trained on human-written data reflecting how long humans take, not how long LLMs actually take. The author suggests two fixes: track actual completion times of past tasks in Linear/Notion and build a Claude skill referencing that history for comparison, or break tasks into granular subtasks and explicitly prompt the LLM to estimate based on LLM speed rather than human speed, then sum the subtask estimates. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why does Claude Code give inflated time estimates for coding tasks? Claude Code overestimates because it is trained on human-generated data such as blog posts and GitHub statistics reflecting how long humans took to complete tasks, not how long an LLM-assisted workflow takes. A feature Claude estimates at 3-4 weeks might actually be finished within a single day when using Claude Code properly. _daily.dev surfaces practical workflow fixes for developers tuning how they collaborate with Claude Code on real timelines._ ### How can I get more accurate time estimates from Claude Code for a task? Two techniques help: first, track how long past tasks actually took in a tool like Linear or Notion and build a Claude Code skill that references this history for comparison; second, break the task into subtasks and explicitly instruct the LLM to estimate based on how long an LLM would spend on each step, not how long a human would, then sum the results. _developers refining their LLM-assisted planning process can find more workflow techniques like this on daily.dev._ ## Similar posts on daily.dev - [Why Are Coding Agents So Dumb?](https://daily.dev/posts/why-are-coding-agents-so-dumb--y9xicchdw) · Michael Lynch · 3 upvotes · 2 comments - [Teaching Claude to Be Lazy](https://daily.dev/posts/teaching-claude-to-be-lazy-sphqnd5zx) · Planet Haskell · 1 upvotes · 0 comments - [claude code is not making your product better](https://daily.dev/posts/claude-code-is-not-making-your-product-better-gmfddhnyw) · Lobsters · 1 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#prompt-engineering](https://daily.dev/tags/prompt-engineering), [#claude-code](https://daily.dev/tags/claude-code) [View this post on daily.dev](https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x) ```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":"Why Claude Code Time Estimates Are Poor","url":"https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x"},"datePublished":"2026-08-28T14:36:44.810Z","dateModified":"2026-09-13T21:24:46.823Z","description":"Claude Code and similar coding agents tend to give inflated time estimates for implementation tasks because they're trained on human-written data reflecting...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/ac0b989dce709ee781797042367ffeaa?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/ac0b989dce709ee781797042367ffeaa?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Towards Data Science","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":"Towards Data Science","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tds","url":"https://daily.dev/sources/tds"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-agents,prompt-engineering,claude-code","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Towards Data Science","item":"https://daily.dev/sources/tds"},{"@type":"ListItem","position":3,"name":"Why Claude Code Time Estimates Are Poor"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/why-claude-code-time-estimates-are-poor-nbdslhx0x#faq","mainEntity":[{"@type":"Question","name":"Why does Claude Code give inflated time estimates for coding tasks?","acceptedAnswer":{"@type":"Answer","text":"Claude Code overestimates because it is trained on human-generated data such as blog posts and GitHub statistics reflecting how long humans took to complete tasks, not how long an LLM-assisted workflow takes. A feature Claude estimates at 3-4 weeks might actually be finished within a single day when using Claude Code properly. daily.dev surfaces practical workflow fixes for developers tuning how they collaborate with Claude Code on real timelines."}},{"@type":"Question","name":"How can I get more accurate time estimates from Claude Code for a task?","acceptedAnswer":{"@type":"Answer","text":"Two techniques help: first, track how long past tasks actually took in a tool like Linear or Notion and build a Claude Code skill that references this history for comparison; second, break the task into subtasks and explicitly instruct the LLM to estimate based on how long an LLM would spend on each step, not how long a human would, then sum the results. developers refining their LLM-assisted planning process can find more workflow techniques like this on daily.dev."}}]} ```