--- title: A mysterious new model just took over the internet... description: A mysterious stealth model called Ox Alpha appeared anonymously on OpenRouter, quickly becoming the most-used model on the platform with 42 trillion tokens... canonical: https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: A mysterious new model just took over the internet... | daily.dev og:description: A mysterious stealth model called Ox Alpha appeared anonymously on OpenRouter, quickly becoming the most-used model on the platform with 42 trillion tokens... og:url: https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn og:image: https://api.daily.dev/og/posts/tFTlDphGn.png og:image:alt: A mysterious new model just took over the internet... 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. # A mysterious new model just took over the internet... **[Fireship](https://daily.dev/sources/fireship)** · 5 min read · 35 upvotes · 2 comments ## Summary A mysterious stealth model called Ox Alpha appeared anonymously on OpenRouter, quickly becoming the most-used model on the platform with 42 trillion tokens served in 6 days, a million-token context window, and video input support. Speculation swirled about its origin until Zhipu AI revealed on August 26th that it was actually GLM 5.3 Flash, a 320-billion-parameter multimodal mixture-of-experts model trained on 100,000 Chinese-made chips, with MIT-licensed weights released on Hugging Face and pricing around 40x cheaper than Claude. The model is tested by rebuilding a decade-old AngularJS side project, showing strong vision and coding capabilities despite being slow and verbose. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What was the mysterious Ox Alpha model on OpenRouter and who made it? Ox Alpha was a stealth release of GLM-5.3 Flash from the Chinese AI lab Zhipu, revealed on August 26th. It is a natively multimodal mixture-of-experts model with 320 billion total parameters, and its MIT-licensed weights were published on Hugging Face the same day the identity was confirmed. _Developers tracking new frontier model releases like GLM 5.3 Flash can follow launches as they unfold on daily.dev._ ### How much does GLM 5.3 Flash cost compared to Claude? GLM 5.3 Flash costs 15 cents per million input tokens and 50 cents per million output tokens, with a 50% discount available through September 9th. That pricing makes it up to 40 times cheaper than Claude, while still holding up reasonably well against independent benchmark scrutiny despite being slow and prone to over-explaining or doom loops. _Anyone comparing model costs before choosing an API can weigh options like GLM 5.3 Flash on daily.dev._ ### How many chips were used to train or serve GLM 5.3 Flash? The stealth run of GLM 5.3 Flash was served on just 100,000 Chinese-made chips, a detail notable because it suggests competitive AI infrastructure can be built without relying on Nvidia hardware. This came as the model's true identity was revealed following its viral anonymous debut as 'Ox Alpha' on OpenRouter. _Following chip and infrastructure shifts behind major model releases is easier with daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@thierryrenematos** · 3 upvotes > its GLM 5.3 flash --- Tags: [#llm](https://daily.dev/tags/llm), [#openrouter](https://daily.dev/tags/openrouter) [View this post on daily.dev](https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn) ```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":"A mysterious new model just took over the internet...","url":"https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn"},"datePublished":"2026-09-01T17:34:03.157Z","dateModified":"2026-09-13T19:34:44.014Z","description":"A mysterious stealth model called Ox Alpha appeared anonymously on OpenRouter, quickly becoming the most-used model on the platform with 42 trillion tokens...","image":"https://i.ytimg.com/vi/r-tzcMlQISk/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/r-tzcMlQISk/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Fireship","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":"Fireship","logo":"https://media.daily.dev/image/upload/s--75ndqrkr--/f_auto,t_logo/v1702882094/logos/fireship.jpg","url":"https://daily.dev/sources/fireship"},"commentCount":2,"discussionUrl":"https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":35},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":2}],"keywords":"llm,openrouter","timeRequired":"PT5M","video":{"@type":"VideoObject","name":"A mysterious new model just took over the internet...","description":"A mysterious stealth model called Ox Alpha appeared anonymously on OpenRouter, quickly becoming the most-used model on the platform with 42 trillion tokens...","thumbnailUrl":"https://i.ytimg.com/vi/r-tzcMlQISk/sddefault.jpg","uploadDate":"2026-09-01T17:34:03.157Z","duration":"PT5M","url":"https://api.daily.dev/r/tFTlDphGn","embedUrl":"https://www.youtube.com/embed/r-tzcMlQISk"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Fireship","item":"https://daily.dev/sources/fireship"},{"@type":"ListItem","position":3,"name":"A mysterious new model just took over the internet..."}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn","comment":[{"@type":"Comment","text":"its GLM 5.3 flash","datePublished":"2026-09-02T20:38:15.959Z","url":"https://daily.dev/posts/tFTlDphGn#c-FJlpqm6xu","author":{"@type":"Person","name":"thierry","url":"https://daily.dev/thierryrenematos","image":"https://avatars.githubusercontent.com/u/1225308?v=4"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/a-mysterious-new-model-just-took-over-the-internet--tftldphgn#faq","mainEntity":[{"@type":"Question","name":"What was the mysterious Ox Alpha model on OpenRouter and who made it?","acceptedAnswer":{"@type":"Answer","text":"Ox Alpha was a stealth release of GLM-5.3 Flash from the Chinese AI lab Zhipu, revealed on August 26th. It is a natively multimodal mixture-of-experts model with 320 billion total parameters, and its MIT-licensed weights were published on Hugging Face the same day the identity was confirmed. Developers tracking new frontier model releases like GLM 5.3 Flash can follow launches as they unfold on daily.dev."}},{"@type":"Question","name":"How much does GLM 5.3 Flash cost compared to Claude?","acceptedAnswer":{"@type":"Answer","text":"GLM 5.3 Flash costs 15 cents per million input tokens and 50 cents per million output tokens, with a 50% discount available through September 9th. That pricing makes it up to 40 times cheaper than Claude, while still holding up reasonably well against independent benchmark scrutiny despite being slow and prone to over-explaining or doom loops. Anyone comparing model costs before choosing an API can weigh options like GLM 5.3 Flash on daily.dev."}},{"@type":"Question","name":"How many chips were used to train or serve GLM 5.3 Flash?","acceptedAnswer":{"@type":"Answer","text":"The stealth run of GLM 5.3 Flash was served on just 100,000 Chinese-made chips, a detail notable because it suggests competitive AI infrastructure can be built without relying on Nvidia hardware. This came as the model's true identity was revealed following its viral anonymous debut as 'Ox Alpha' on OpenRouter. Following chip and infrastructure shifts behind major model releases is easier with daily.dev."}}]} ```