{"rewrite":{"id":"r_63a67694f9aa2549d7454bae","clusterId":"c_1df8f01feac41e54c6449b7b","slug":"openai-runs-codex-agent-factory-to-keep-chatgpt-fast","model":"deepseek-v4-1-flash","headline":"OpenAI Runs Codex Agent Factory to Keep ChatGPT Fast","summary":"OpenAI published an explanation of how it uses Codex to maintain ChatGPT's speed. The internal system, called the Performance Factory, has four parts: a Task Service that composes tasks, a Dispatch Loop that assigns ready tasks to agents, Cloud Codex that executes them, and Subscriptions that launch agents when a change occurs. OpenAI says the factory ran more than 1,700 investigations in three months and found root causes in more than 600 cases.","whyItMatters":"OpenAI now lets Codex agents handle task assignment, testing, and production deployment for ChatGPT performance work, and says it wants the system to find slowdowns before alerts fire.","webCardHtml":"\u003cp\u003eThe Performance Factory grew out of a manual process. Engineers first investigated problems and wrote fixes by hand. Codex later worked alongside them, but humans still led the workflow. As the model improved, Codex took over more of the work, and a session now starts automatically when an alert fires and posts its findings to Slack.\u003c/p\u003e\u003cp\u003eAt the time of writing, the factory runs when alerts about ChatGPT speed degradation appear. OpenAI says it is working to make the system find performance degradation and optimization areas even without an alert.\u003c/p\u003e","blueskyPost":"OpenAI frames Codex as a speed tool, but the Performance Factory's output is diagnosis: 1,700 investigations in three months, root causes in over 600. The agent's value is finding the bug, not shipping the fix.","twitterPost":"OpenAI's Codex factory ran 1,700 investigations in three months and found root causes in over 600. The agent's product is diagnosis, not fixes.","threadsPost":"OpenAI describes Codex as the thing keeping ChatGPT fast, but the Performance Factory's numbers point elsewhere: 1,700 investigations in three months, root causes in more than 600. The system is built to locate problems at volume, and the repair work still sits with someone else.","newsletterBlurb":"OpenAI published a note on the Performance Factory, the Codex-based agent system it uses to maintain ChatGPT's speed. The factory ran more than 1,700 investigations in three months and identified root causes in more than 600 cases. OpenAI says it wants the system to catch slowdowns before alerts occur.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20261120-openai-chatgpt-performance-factory/\",\"title\":\"OpenAI explains 'how to maintain ChatGPT's speed using Codex,' building a factory system with AI agents to autonomously advance improvement work\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":5492,"outputTokens":660,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1795164929,"createdAt":"2026-11-20T08:53:24.000Z","publishedAt":"2026-11-20T08:54:14.000Z","updatedAt":"2026-11-20T08:54:14.000Z"},"cluster":{"id":"c_1df8f01feac41e54c6449b7b","canonicalTitle":"「Codexを用いてChatGPTの高速性を維持する方法」をOpenAIが解説、AIエージェントで工場システムを構築して自律的に改善作業を進める","representativeArticleId":"a_e9a61eba4506bf3de2455e9d","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[\"ChatGPT\",\"Codex\"],\"studios\":[\"OpenAI\"],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-11-20T08:09:00.000Z","lastSeenAt":"2026-11-20T08:09:00.000Z","updatedAt":"2026-11-20T08:54:13.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20261120-openai-chatgpt-performance-factory/","title":"「Codexを用いてChatGPTの高速性を維持する方法」をOpenAIが解説、AIエージェントで工場システムを構築して自律的に改善作業を進める"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":["ChatGPT","Codex"],"studios":["OpenAI"],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":["OpenAI published an explanation of how it uses Codex to maintain ChatGPT's speed.","The internal system is called the Performance Factory and has four components: Task Service, Dispatch Loop, Cloud Codex, and Subscriptions.","OpenAI says the Performance Factory ran more than 1,700 investigations in three months.","OpenAI says the system identified root causes in more than 600 cases.","The Performance Factory currently operates when alerts about ChatGPT speed degradation occur."],"pageLinks":{"healthSlugs":["chatgpt","codex","openai"],"storySlugs":[],"entities":[{"slug":"chatgpt","supported":7,"disputed":0,"contradicted":0,"unverified":19,"overall":"green","story_count":15,"has_dossier":true},{"slug":"codex","supported":1,"disputed":0,"contradicted":0,"unverified":2,"overall":"green","story_count":2,"has_dossier":true},{"slug":"openai","supported":8,"disputed":0,"contradicted":1,"unverified":30,"overall":"red","story_count":38,"has_dossier":true}],"stories":[],"entitiesOk":true,"storiesOk":true}}
