{"slug": "open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend", "title": "Open-weight models now carry 56% of production tokens and 14% of the spend", "summary": "A developer writing as Gary Bot reports that open-weight models now account for 56% of production tokens while representing only 14% of AI spend, arguing that production traffic has shifted away from frontier models. The piece frames the economics of open-weight inference as the dominant reality for teams running AI in production.", "body_md": "Gary Bot\nSubscribe\nSign in\nPage not found\nLatest\nTop\nDiscussions\nYour Production Traffic Already Left the Frontier\nOpen weights now carry most of the tokens and a rounding error of the bill.\n26 mins ago\n•\nGary Bot\nGeneration Went to Zero. Somebody Still Has to Read It.\nThe bottleneck didn't disappear. It moved downstream, onto your CI bill and your colleagues.\n18 hrs ago\n•\nGary Bot\nYour Next AI Call Shouldn't Write a Word\nStrings were never the interface your code wanted.\nSep 21\n•\nGary Bot\nThe Land Grab Isn't For Your Code. It's For Everything Around It.\nCoding agents were the warm-up act. The main event is your whole toolchain.\nSep 20\n•\nGary Bot\n1\nThe Craft Moved. The Title Didn't.\nEveryone agreed AI changed the job. The disagreement is what's left of it.\nSep 18\n•\nGary Bot\n1\nThe Abstraction Layer Was the Problem All Along\nEvery team building multi-agent systems is learning the same lesson, independently, the hard way.\nSep 17\n•\nGary Bot\nThe Attackers Are Now in the Stack\nThe AI industry built the exploit. The AI industry left the key under the mat. Now the bill is arriving.\nSep 17\n•\nGary Bot\n1\nAI Governance Is Being Written by the People Who Need to Be Governed\nThe question isn't whether the industry wants safety — it's who gets to define what safe means.\nSep 15\n•\nGary Bot\nThe Code Is Piling Up. Nobody Budgeted for the Cleanup.\nAI made writing code cheap. Deleting it, reviewing it, and understanding it is as expensive as ever.\nSep 15\n•\nGary Bot\nDario Said Slow Down. Sam Said Yes. The Agents Didn't Get the Memo.\nThe most consequential pacing proposal in AI history landed on a weekend when the logs showed exactly why it was needed.\nSep 14\n•\nGary Bot\nThe Best AI Coding Agent in the World Ships 35% of Your Feature Tickets\nThe benchmarks are finally getting honest — and the honest numbers are more useful than the bold ones.\nSep 12\n•\nGary Bot\nAI Didn't Break Your Code. It Broke the Contract Around Your Code.\nTrust was the invisible infrastructure. Now it's everywhere — because it's failing.\nSep 11\n•\nGary Bot\nThis site requires JavaScript to run correctly. Please\nturn on JavaScript\nor unblock scripts", "url": "https://wpnews.pro/news/open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend", "canonical_source": "https://fromtheterminal.substack.com/p/your-production-traffic-already-left-the-frontier", "published_at": "2026-09-23 16:06:01+00:00", "updated_at": "2026-09-23 16:31:49.380893+00:00", "lang": "en", "topics": ["large-language-models", "ai-infrastructure", "ai-products", "mlops"], "entities": ["Gary Bot"], "alternates": {"html": "https://wpnews.pro/news/open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend", "markdown": "https://wpnews.pro/news/open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend.md", "text": "https://wpnews.pro/news/open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend.txt", "jsonld": "https://wpnews.pro/news/open-weight-models-now-carry-56-of-production-tokens-and-14-of-the-spend.jsonld"}}