--- title: Stop building agents that can’t change models | daily.dev description: Agents that treat one AI model as their foundation break when that provider changes its offerings, so the fix is to design a stable contract between agents and... canonical: https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Stop building agents that can’t change models | daily.dev og:description: Agents that treat one AI model as their foundation break when that provider changes its offerings, so the fix is to design a stable contract between agents and... og:url: https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya og:image: https://api.daily.dev/og/posts/abqvQk2YA.png og:image:alt: Stop building agents that can’t change models 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. # Stop building agents that can’t change models **[InfoWorld](https://daily.dev/sources/infoworld)** · 7 min read · 0 upvotes · 0 comments ## Summary Agents that treat one AI model as their foundation break when that provider changes its offerings, so the fix is to design a stable contract between agents and models. OpenAI's enterprise LLM API share fell from 50% in 2023 to 27% by the end of 2025 while Anthropic rose from 12% to 40%. Anthropic retired Claude Sonnet 4 and Opus 4 on June 15, 2026, and OpenAI shut down its Assistants API. Only 11% of teams switched providers in a year per Menlo's mid-2025 survey, while 66% upgraded within their existing vendor. Keeping provider SDKs, state, prompts and fine-tunes out of business logic, plus an independent evaluation suite, makes switching a configuration change followed by an evaluation run instead of a rebuild. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What happened when Anthropic retired Claude Sonnet 4 and Opus 4 from its API? Anthropic retired Claude Sonnet 4 and Claude Opus 4 from its API on June 15, 2026, after giving developers notice on April 14. Any application still calling those model IDs started receiving errors. Claude Opus 4.1 followed on August 5, and Anthropic commits to at least 60 days of notice for publicly released models going forward. _Teams planning model migrations can track provider deprecation notices and architecture patterns on daily.dev._ ### How has enterprise LLM API market share shifted between OpenAI, Anthropic, and Google? OpenAI's share of enterprise LLM API usage fell from 50% in 2023 to 27% by the end of 2025, according to Menlo Ventures. Anthropic grew from 12% to 40% in the same period, and Google rose from 7% to 21%, showing leadership in the model market changes faster than most enterprise software contracts renew. _Developers weighing which LLM provider to bet on can follow market shifts like this on daily.dev._ ### How should I structure an AI agent so I can switch LLM providers without rewriting it? Build a stable interface layer so the agent never imports a provider SDK directly: define a ModelClient protocol with a complete() method taking a ModelRequest and returning a ModelResponse, use plain JSON Schema for tools, and map each task to an adapter via configuration. Keep state, embeddings, and fine-tunes outside the provider, and maintain an independent evaluation suite to validate replacements. _Engineers designing provider-agnostic agent architectures can find patterns like this on daily.dev._ ## Similar posts on daily.dev - [Mistral Says 99% of Enterprise AI Can Be Open. That Should Terrify the Closed-Model Business](https://daily.dev/posts/mistral-says-99-of-enterprise-ai-can-be-open-that-should-terrify-the-closed-model-business-eq4tuzn9x) · Medium · 1 upvotes · 0 comments - [Weekly Dose \#4 - From Smarter Models to Safer Systems](https://daily.dev/posts/weekly-dose-4---from-smarter-models-to-safer-systems-yqez1xpcq) · Machine Learning Pills · 1 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#llm](https://daily.dev/tags/llm), [#architecture](https://daily.dev/tags/architecture), [#mcp](https://daily.dev/tags/mcp), [#anthropic](https://daily.dev/tags/anthropic) [View this post on daily.dev](https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya) ```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":"Stop building agents that can’t change models","url":"https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya"},"datePublished":"2026-10-26T09:04:55.623Z","dateModified":"2026-10-26T10:33:26.994Z","description":"Agents that treat one AI model as their foundation break when that provider changes its offerings, so the fix is to design a stable contract between agents and...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f184eb95756ad2ac353b8ce5b45eeb17?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f184eb95756ad2ac353b8ce5b45eeb17?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfoWorld","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":"InfoWorld","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/bf6d68a999064029b0bb09aa6268f1f3","url":"https://daily.dev/sources/infoworld"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,llm,architecture,mcp,anthropic","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfoWorld","item":"https://daily.dev/sources/infoworld"},{"@type":"ListItem","position":3,"name":"Stop building agents that can’t change models"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/stop-building-agents-that-can-t-change-models-abqvqk2ya#faq","mainEntity":[{"@type":"Question","name":"What happened when Anthropic retired Claude Sonnet 4 and Opus 4 from its API?","acceptedAnswer":{"@type":"Answer","text":"Anthropic retired Claude Sonnet 4 and Claude Opus 4 from its API on June 15, 2026, after giving developers notice on April 14. Any application still calling those model IDs started receiving errors. Claude Opus 4.1 followed on August 5, and Anthropic commits to at least 60 days of notice for publicly released models going forward. Teams planning model migrations can track provider deprecation notices and architecture patterns on daily.dev."}},{"@type":"Question","name":"How has enterprise LLM API market share shifted between OpenAI, Anthropic, and Google?","acceptedAnswer":{"@type":"Answer","text":"OpenAI's share of enterprise LLM API usage fell from 50% in 2023 to 27% by the end of 2025, according to Menlo Ventures. Anthropic grew from 12% to 40% in the same period, and Google rose from 7% to 21%, showing leadership in the model market changes faster than most enterprise software contracts renew. Developers weighing which LLM provider to bet on can follow market shifts like this on daily.dev."}},{"@type":"Question","name":"How should I structure an AI agent so I can switch LLM providers without rewriting it?","acceptedAnswer":{"@type":"Answer","text":"Build a stable interface layer so the agent never imports a provider SDK directly: define a ModelClient protocol with a complete() method taking a ModelRequest and returning a ModelResponse, use plain JSON Schema for tools, and map each task to an adapter via configuration. Keep state, embeddings, and fine-tunes outside the provider, and maintain an independent evaluation suite to validate replacements. Engineers designing provider-agnostic agent architectures can find patterns like this on daily.dev."}}]} ```