--- title: The AI Agent Ate My Budget! What I Should Have Built Instead description: An opinion piece argues that AI agents which reason at runtime are becoming financially risky as token costs and usage climb, and that deterministic... canonical: https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The AI Agent Ate My Budget! What I Should Have Built Instead | daily.dev og:description: An opinion piece argues that AI agents which reason at runtime are becoming financially risky as token costs and usage climb, and that deterministic... og:url: https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci og:image: https://api.daily.dev/og/posts/0bfaiwQCI.png og:image:alt: The AI Agent Ate My Budget! What I Should Have Built Instead 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. # The AI Agent Ate My Budget! What I Should Have Built Instead **[Telerik](https://daily.dev/sources/telerik)** · 6 min read · 0 upvotes · 0 comments ## Summary An opinion piece argues that AI agents which reason at runtime are becoming financially risky as token costs and usage climb, and that deterministic automation—systems built on fixed rules rather than runtime reasoning—offers a more predictable, cost-controlled alternative. The author describes building a code generator that produces a full production-grade SaaS stack (backend, frontend, tests, observability) deterministically in under eight hours with zero tokens, versus using an AI coding assistant for test generation at a controlled $20 cost. The piece advocates using AI surgically with spending limits rather than as a default execution engine, and mentions exploring local LLMs via Ollama to avoid dependency on external AI pricing. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the difference between an AI agent and deterministic automation in software development? An AI agent receives a goal, reasons about it, and makes runtime decisions, consuming tokens with every step, making its cost unpredictable. Deterministic automation analyzes a known structure and applies fixed rules to produce predictable output without tokens or runtime reasoning, since the reasoning was front-loaded into the system's design rather than executed live. _Developers weighing agents against fixed automation can track real-world cost tradeoffs on daily.dev._ ### How much does it cost to use an AI coding assistant to generate unit tests for an existing codebase? One developer reported spending around $20 in tokens to expand unit test coverage using an AI coding assistant, running it for about six hours over two days against an existing test suite with consistent patterns. The cost stayed low because a spending limit was set in advance; without that cap, the same task could have spiraled into hundreds or thousands of dollars. _Anyone budgeting AI-assisted coding work can compare cost-control approaches on daily.dev._ ### What hardware do I need to run a custom LLM locally with Ollama for coding assistance? Running a custom LLM locally with Ollama for coding integration demands serious hardware, specifically a GPU with at least 16 to 20 GB of VRAM. In exchange, it offers unlimited token usage at zero marginal cost, decoupling productivity from external providers' pricing models, though it requires upfront infrastructure investment. _Teams considering self-hosted LLMs for coding can weigh the hardware tradeoffs via daily.dev._ ## Similar posts on daily.dev - [AI's Economics Don't Make Sense](https://daily.dev/posts/ai-s-economics-don-t-make-sense-u59jqqbul) · Where's Your Ed At · 69 upvotes · 11 comments - [Microsoft’s quiet Claude Code retreat and the real cost of enterprise AI](https://daily.dev/posts/microsoft-s-quiet-claude-code-retreat-and-the-real-cost-of-enterprise-ai-8ycsrgrmm) · The Next Web · 44 upvotes · 6 comments - [The real cost of agentic AI](https://daily.dev/posts/the-real-cost-of-agentic-ai-q32iamekb) · InfoWorld · 0 upvotes · 0 comments - [Sticker shock has execs rethinking this whole AI thing](https://daily.dev/posts/sticker-shock-has-execs-rethinking-this-whole-ai-thing-ywgulwvsx) · The Register · 0 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#ollama](https://daily.dev/tags/ollama) [View this post on daily.dev](https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci) ```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":"The AI Agent Ate My Budget! What I Should Have Built Instead","url":"https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci"},"datePublished":"2026-08-17T18:10:31.650Z","dateModified":"2026-08-17T18:12:51.763Z","description":"An opinion piece argues that AI agents which reason at runtime are becoming financially risky as token costs and usage climb, and that deterministic...","image":"https://media.daily.dev/image/upload/s--0_ODbtD2--/f_auto/v1722860399/public/Placeholder%2008","thumbnailUrl":"https://media.daily.dev/image/upload/s--0_ODbtD2--/f_auto/v1722860399/public/Placeholder%2008","isAccessibleForFree":true,"articleSection":"Telerik","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":"Telerik","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/5ee6d4bef8ac4f84886d06f5881963e7","url":"https://daily.dev/sources/telerik"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,ollama","timeRequired":"PT6M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Telerik","item":"https://daily.dev/sources/telerik"},{"@type":"ListItem","position":3,"name":"The AI Agent Ate My Budget! What I Should Have Built Instead"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-ai-agent-ate-my-budget-what-i-should-have-built-instead-0bfaiwqci#faq","mainEntity":[{"@type":"Question","name":"What is the difference between an AI agent and deterministic automation in software development?","acceptedAnswer":{"@type":"Answer","text":"An AI agent receives a goal, reasons about it, and makes runtime decisions, consuming tokens with every step, making its cost unpredictable. Deterministic automation analyzes a known structure and applies fixed rules to produce predictable output without tokens or runtime reasoning, since the reasoning was front-loaded into the system's design rather than executed live. Developers weighing agents against fixed automation can track real-world cost tradeoffs on daily.dev."}},{"@type":"Question","name":"How much does it cost to use an AI coding assistant to generate unit tests for an existing codebase?","acceptedAnswer":{"@type":"Answer","text":"One developer reported spending around $20 in tokens to expand unit test coverage using an AI coding assistant, running it for about six hours over two days against an existing test suite with consistent patterns. The cost stayed low because a spending limit was set in advance; without that cap, the same task could have spiraled into hundreds or thousands of dollars. Anyone budgeting AI-assisted coding work can compare cost-control approaches on daily.dev."}},{"@type":"Question","name":"What hardware do I need to run a custom LLM locally with Ollama for coding assistance?","acceptedAnswer":{"@type":"Answer","text":"Running a custom LLM locally with Ollama for coding integration demands serious hardware, specifically a GPU with at least 16 to 20 GB of VRAM. In exchange, it offers unlimited token usage at zero marginal cost, decoupling productivity from external providers' pricing models, though it requires upfront infrastructure investment. Teams considering self-hosted LLMs for coding can weigh the hardware tradeoffs via daily.dev."}}]} ```