--- title: How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY... description: A practical guide to maximizing Kimi K3's capabilities through its full subscription plan, proper prompting strategies, and the CLI's goals system. The plan... canonical: https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS! | daily.dev og:description: A practical guide to maximizing Kimi K3's capabilities through its full subscription plan, proper prompting strategies, and the CLI's goals system. The plan... og:url: https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz og:image: https://api.daily.dev/og/posts/azfI6aSiz.png og:image:alt: How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS! 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. # How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS! **[AICodeKing](https://daily.dev/sources/aicodeking)** · 14 min read · 0 upvotes · 0 comments ## Summary A practical guide to maximizing Kimi K3's capabilities through its full subscription plan, proper prompting strategies, and the CLI's goals system. The plan includes Kimi CLI, a web app, VS Code extension, Kimi Claw (cloud-hosted Open Claw deployment), and Kimi Work (local desktop agent with 300-agent swarm). Key insight: K3 performs best when prompted toward a final goal rather than step-by-step instructions. The goals system (/goal command) lets users declare an end state the agent works toward autonomously across multiple turns, supports queuing, headless/CI mode with exit codes, and a 'write goal' feature that auto-expands simple prompts into structured objectives. Additional CLI features covered include mid-turn steering, /btw side conversations, plan mode, swarm parallelism, and scheduled cron tasks. ## 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 use my Kimi coding plan API key inside Claude Code? Route it through the open-source CLI Proxy API tool, which wraps providers behind local OpenAI, Gemini, and Claude compatible endpoints. Add a Claude API key entry in the proxy's config.yaml with your SK-Kimi key, set the base URL to api.kimi.com/coding, and add a user-agent header identifying it as Kimi CLI so the endpoint accepts requests. Then point Claude Code's Anthropic base URL at localhost:8317. _daily.dev surfaces workflow tricks like this for developers wiring one model's quota into another tool._ ### How should I prompt Kimi K3 differently from other coding models? Describe the final goal and how to verify it rather than breaking the task into small steps fed one at a time. K3 doesn't give up or stall on long tasks, so interrupting it with incremental instructions actually hurts performance; instead state the finish line, like full working authentication with validation and passing tests, and let it run autonomously. _developers tuning prompts for long-horizon agents often compare notes on daily.dev before rewriting their workflow._ ### What is the goal mode feature in Kimi CLI and how does it work? Goal mode lets a developer declare an end state with the /goal command instead of giving step-by-step instructions, and the agent checks after every turn whether the goal is complete, blocked, or still active until it finishes. Goals support pausing, resuming, queuing multiple objectives, turn and wall-clock budgets, and headless execution where exit code 0 means completed, 3 means blocked, and 6 means paused, enabling use inside CI pipelines. _engineers scripting autonomous coding agents into CI track feature deep-dives like this via daily.dev._ ## Similar posts on daily.dev - [Introducing enhancements in Kiro CLI 2.0](https://daily.dev/posts/introducing-enhancements-in-kiro-cli-2-0-r0tzwdukt) · SD Times · 38 upvotes · 1 comments - [I Tried Kimi AI in 2026: 10 Things It Can Do That Surprised Me](https://daily.dev/posts/i-tried-kimi-ai-in-2026-10-things-it-can-do-that-surprised-me-mjsgk4kez) · Medium · 0 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#claude-code](https://daily.dev/tags/claude-code), [#kimi](https://daily.dev/tags/kimi) [View this post on daily.dev](https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz) ```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":"How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS!","url":"https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz"},"datePublished":"2026-07-17T09:28:13.928Z","dateModified":"2026-09-14T06:38:38.597Z","description":"A practical guide to maximizing Kimi K3's capabilities through its full subscription plan, proper prompting strategies, and the CLI's goals system. The plan...","image":"https://i.ytimg.com/vi/Hh9aWyTmnyQ/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/Hh9aWyTmnyQ/sddefault.jpg","isAccessibleForFree":true,"articleSection":"AICodeKing","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":"AICodeKing","logo":"https://media.daily.dev/image/upload/s--x7nDUfWj--/f_auto,q_auto/v1768208312/logos/aicodeking","url":"https://daily.dev/sources/aicodeking"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,claude-code,kimi","timeRequired":"PT14M","video":{"@type":"VideoObject","name":"How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS!","description":"A practical guide to maximizing Kimi K3's capabilities through its full subscription plan, proper prompting strategies, and the CLI's goals system. The plan...","thumbnailUrl":"https://i.ytimg.com/vi/Hh9aWyTmnyQ/sddefault.jpg","uploadDate":"2026-07-17T09:28:13.928Z","duration":"PT14M","url":"https://api.daily.dev/r/azfI6aSiz","embedUrl":"https://www.youtube.com/embed/Hh9aWyTmnyQ"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"AICodeKing","item":"https://daily.dev/sources/aicodeking"},{"@type":"ListItem","position":3,"name":"How to USE Kimi K3 to it's EXTREME: I RAN IT CONTINUOSLY FOR 48HRS!"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/how-to-use-kimi-k3-to-it-s-extreme-i-ran-it-continuosly-for-48hrs--azfi6asiz#faq","mainEntity":[{"@type":"Question","name":"How do I use my Kimi coding plan API key inside Claude Code?","acceptedAnswer":{"@type":"Answer","text":"Route it through the open-source CLI Proxy API tool, which wraps providers behind local OpenAI, Gemini, and Claude compatible endpoints. Add a Claude API key entry in the proxy's config.yaml with your SK-Kimi key, set the base URL to api.kimi.com/coding, and add a user-agent header identifying it as Kimi CLI so the endpoint accepts requests. Then point Claude Code's Anthropic base URL at localhost:8317. daily.dev surfaces workflow tricks like this for developers wiring one model's quota into another tool."}},{"@type":"Question","name":"How should I prompt Kimi K3 differently from other coding models?","acceptedAnswer":{"@type":"Answer","text":"Describe the final goal and how to verify it rather than breaking the task into small steps fed one at a time. K3 doesn't give up or stall on long tasks, so interrupting it with incremental instructions actually hurts performance; instead state the finish line, like full working authentication with validation and passing tests, and let it run autonomously. developers tuning prompts for long-horizon agents often compare notes on daily.dev before rewriting their workflow."}},{"@type":"Question","name":"What is the goal mode feature in Kimi CLI and how does it work?","acceptedAnswer":{"@type":"Answer","text":"Goal mode lets a developer declare an end state with the /goal command instead of giving step-by-step instructions, and the agent checks after every turn whether the goal is complete, blocked, or still active until it finishes. Goals support pausing, resuming, queuing multiple objectives, turn and wall-clock budgets, and headless execution where exit code 0 means completed, 3 means blocked, and 6 means paused, enabling use inside CI pipelines. engineers scripting autonomous coding agents into CI track feature deep-dives like this via daily.dev."}}]} ```