{"slug":"z-ai-alpha","label":"Z.ai Alpha","item_count":24,"day_count":9,"source_count":2,"first_seen":"2026-08-12T05:18:40+00:00","last_updated":"2026-08-31T10:08:21+00:00","via_scout":true,"generated_at":"2026-09-02T05:08:18.069258+00:00","sources":["hackernews_ai","search_cn_open_weight_labs"],"days":[{"date":"2026-08-12","items":[{"title":"Zhipu’s API User Base Nears 7 Million as It Adds 50,000-Plus Chinese AI Chips - TechNode","url":"https://news.google.com/rss/articles/CBMirAFBVV95cUxQMUVtdXlPeGI2b3JhWk56UDZTMk1UUmU0aktzR1F6OENrY19PdTdYY2xZbXROUTBER0Nwd1FJQkRLTWZnREtaSk4xaDFpTkJOajZIVzNjQmduVUR0eHVZNHg0bnNtbVpHNjNDYmhHZTVyYjBtZEVBRWstc2gweUVlTDdyQms3dnh6cHNPZTg1eEs3Tk1IOWlveTZTSmtTU2t1b2JZLVlDS3ZPU05T?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Zhipu’s API User Base Nears 7 Million as It Adds 50,000-Plus Chinese AI Chips TechNode","sid":"fde0e6704e2e8662","published":"2026-08-12T05:18:40+00:00","editor_note":"TechNode reports Zhipu's API user base nears 7 million as it quietly adds over 50,000 Chinese AI chips — weeks before that chip choice becomes the Ox Alpha story."}]},{"date":"2026-08-14","items":[{"title":"Zhipu AI Releases GLM-5.3: Coding Capability Jumps 50%, Open-Source Weights Coming in Two Weeks - finance.biggo.com","url":"https://news.google.com/rss/articles/CBMidkFVX3lxTE5kSzl6Snp6UG5tQ3lUU0gtWlhPRm5fVzJJZ0hQNlo0cldzWmk2SHV6enpWZU1COUFCMjNxR0VIcXFuRkpNRUhydjJ4bWVRdHpWS2VBNXpsaU41WU15VlpEMDN1UTdZenVraXpoZGtmdFFZcXhNQ2c?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Zhipu AI Releases GLM-5.3: Coding Capability Jumps 50%, Open-Source Weights Coming in Two Weeks finance.biggo.com","sid":"1359156e75b44e45","published":"2026-08-14T06:05:00+00:00","editor_note":"Zhipu releases GLM-5.3, claiming a 50% coding-capability jump and promising open-source weights within two weeks."},{"title":"Zhipu AI releases GLM-5.3, claims it's the strongest open-weights coding model - the-decoder.com","url":"https://news.google.com/rss/articles/CBMiowFBVV95cUxNM0JXWjNoMDQ1Zl93ZWRmejA4djlQQzVodHhYb0lCQnJqSGtFbUxTUDQyYnI3NmN6NF9aazJRTkx4YlZ6b2ROVzNPQUpKTTgtVGh2bDU1aUE0RXItRUV0RE1tckNPNjhoUkxhNUs3ejBJNVAtZU1LMDR4cllacVBDZjV3UU96TnFMbkxqaUdTa3pacWhneDVUQmpLd1hsMmFrcGlB?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Zhipu AI releases GLM-5.3, claims it's the strongest open-weights coding model the-decoder.com","sid":"842cf30c053df9d2","published":"2026-08-14T10:24:31+00:00","editor_note":"the-decoder.com covers the same GLM-5.3 release, with Zhipu billing it as the strongest open-weights coding model."}]},{"date":"2026-08-18","items":[{"title":"China Closes in on Cyber Race as GLM-5.3 Beats US Models - Cyber Magazine","url":"https://news.google.com/rss/articles/CBMikAFBVV95cUxONW82TWREQ2JhRzllMGZtQnV5RVZHY0xqQVdWV2VGdW0wZkxEMFVVZ0thZ1BoTlpQTkNOSkN5RUUwclRkaVBXMWN0RVN6Y0J2ZkJaWm11VDhEV2tSMnFHc25ONU9MSnpNM1hiTlFSZU11TVFmYWVucWpzbVREdURTWktYZEt5MkhzUjI5RmEtMWs?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"China Closes in on Cyber Race as GLM-5.3 Beats US Models Cyber Magazine","sid":"7216c7465c1c56b5","published":"2026-08-18T07:34:13+00:00","editor_note":"Cyber Magazine reports GLM-5.3 beating US models on cyber-race benchmarks, before the model has open weights."},{"title":"Reading Zhipu’s GLM-5.3 results past the headline number - AI News","url":"https://news.google.com/rss/articles/CBMijAFBVV95cUxQd1FvdG5QWDRCRTNtSG1icmdNWl9BZWRvUzEzaGFLUnl4ZzlvZ1hCekhCRUhUTVhqclRvdEp2enZKTGdmT1NDZjJ5VTJkRlRIMFRBTDN4dkRGYkZzcXQ2dUYyM3l4cWNuOXRXTldwMzY2SThJYlRxMmZwWEJPaDE1UWZfeUFLNFR5ZkJZUg?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Reading Zhipu’s GLM-5.3 results past the headline number AI News","sid":"d209b402e386f93b","published":"2026-08-18T10:00:48+00:00","editor_note":"AI News reads the GLM-5.3 benchmark results past the headline number, the same day as the cyber-race claim."}]},{"date":"2026-08-19","items":[{"title":"Z.ai says GLM-5.3 lifts coding and cyber tests, flags Cursor vulnerability - digitimes","url":"https://news.google.com/rss/articles/CBMijwFBVV95cUxOT2FRb3R6d0tLWXZ3RUdIQmF5bHY3VktkUnhOdmRiT1VhX1lMUS1tTkdIZnFBU0FncTdSQVVQNVhqQmhLTEpmelZjMGlIR1ozUlJZb014clRvVnl1QkJncGk1anFNVEZRRnl0clI1Wkc4N0NDNmJDdS1xeklRb1E5djVsRmZsX19Oc0JfaEJ4cw?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.ai says GLM-5.3 lifts coding and cyber tests, flags Cursor vulnerability digitimes","sid":"6476118b701d8fd0","published":"2026-08-19T00:42:25+00:00","editor_note":"Z.ai's own results add coding and cyber-test gains but flag a vulnerability in Cursor."}]},{"date":"2026-08-26","items":[{"title":"Z.ai confirms Ox Alpha is a new GLM-series model and will release its weights","url":"https://www.bloomberg.com/news/articles/2026-08-26/china-s-z-ai-made-ox-alpha-stealth-model-that-rivals-deepseek","source":"hackernews_ai","type":"release","summary_1line":"Z.ai confirms Ox Alpha is a new GLM-series model and will release its weights","sid":"a9c204e117f936d5","published":"2026-08-26T10:04:55+00:00","editor_note":"Bloomberg reports Z.ai confirmed the unattributed \"Ox Alpha\" model, which had been benchmarked rivaling DeepSeek, is its own upcoming GLM-series release."},{"title":"Z.ai Releases GLM-5.3-Flash: A 320B-A18B Natively Multimodal MoE With a 1M-Token Context","url":"https://news.google.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?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Z.ai Releases GLM-5.3-Flash: A 320B-A18B Natively Multimodal MoE With a 1M-Token Context MarkTechPost","sid":"05d40e9a614e0aad","published":"2026-08-26T21:33:23+00:00","editor_note":"MarkTechPost details the GLM-5.3-Flash release: a 320B-A18B natively multimodal MoE with a 1M-token context."}]},{"date":"2026-08-27","items":[{"title":"The Computing Power Secret Behind Zhipu AI's \"Niu Lai\" Model: 100% Domestic Chip Support, Priced at Only 1/40 of Opus 4.8","url":"https://news.google.com/rss/articles/CBMiU0FVX3lxTFBwSThwYk5pUy1ObXBtYk5hT1VwbUZJMWdUR19rUGl1bzJqbFppX2JKRDN2cG5fSENyT1J6MUtQWFpWV0IxZXhpSGhlM1Z0b250MnFJ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"The Computing Power Secret Behind Zhipu AI's \"Niu Lai\" Model: 100% Domestic Chip Support, Priced at Only 1/40 of Opus 4.8 36 Kr","sid":"5bebd783d4b4b151","published":"2026-08-27T01:35:04+00:00","editor_note":"A Chinese-language deep dive on the \"Niu Lai\" (Ox) build's computing-power economics: 100% domestic chip support priced near 1/40 of Opus 4.8."},{"title":"Z.ai Built Ox Alpha: GLM-5.3-Flash Now Open-Source","url":"https://news.google.com/rss/articles/CBMidEFVX3lxTE16SGoyXzFTZktwbjlRSU5jZWtTaXctLTBtQmE5UExNVE1BdUVNaXc1elc3THRTeGVCSTNBQ3J6UTVQMkJDTUktY1FaX29BaE9mMTE5SHh3OU8tT29DM3VoaU13Ml90WmxyYlBiaWUwenhacmhB?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.ai Built Ox Alpha: GLM-5.3-Flash Now Open-Source Intelligent Living","sid":"135778ccf99d98eb","published":"2026-08-27T02:36:53+00:00","editor_note":"Z.ai identifies Ox Alpha as GLM-5.3-Flash and open-sources the weights."},{"title":"Z.ai shares surge 8% after releasing new AI model running only on Chinese chips","url":"https://news.google.com/rss/articles/CBMijwFBVV95cUxNOXJYOG11aWZmNUM0QmxZXzloZEdQel9qaTFBbmRzc1VFV0ZiUFNOMHp0VjNFd1BpQU56X3E4YzA2S0tINnd2TEhSc19uM19WSG1NUG9kblZkamRFcURkNE81Ty16c1dmMXVTYW9CSG5tOEpFN1ZyRnFZeTNLdktSRHhRNnhpaEo3RUt5OUVXY9IBlAFBVV95cUxNTHZVd2RGQndFcXFYcTFITlJZcGJvZ0VLeDV0ZnFQX1JGRXBTS2tJOUdZdXh1Q0FrdGtZV3BxMGJvU1E2MDg2MnVpYnFWYzZtNTAyY3NKVTNOd0FGTXZXbkJoY3daVzk2RWY2bVIxRnlFTXpyUUpreVFkTXg4bVZtYmRmVU9GNTVuR3B2NFB5M2hHU1Zq?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.ai shares surge 8% after releasing new AI model running only on Chinese chips CNBC","sid":"bc33de06305fc366","published":"2026-08-27T03:20:00+00:00","editor_note":"Z.ai-linked shares surge 8% after the domestic-chip model release goes public."},{"title":"Zhipu AI releases GLM-5.3-Flash model, powered by 100,000 domestically made chips","url":"https://news.google.com/rss/articles/CBMiYkFVX3lxTE1MdlNBdE9pQ0VZMGs5VXJEUERVSkhQNzJZZlI1cVctaUoyTFA4M2VweS01UU1QRHQxUEthNGNBSTd4VlNFVjk4RHN6YjFlbGcyRF8zMGpaMlhiTXB6ckZ0QW53?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Zhipu AI releases GLM-5.3-Flash model, powered by 100,000 domestically made chips Global Times","sid":"7e30319c44f49483","published":"2026-08-27T05:03:00+00:00","editor_note":"Global Times reports the GLM-5.3-Flash release itself is powered by 100,000 domestically made chips, folding the chip angle directly into the release story."},{"title":"Zhipu identifies Ox Alpha as GLM-5.3-Flash and releases model weights","url":"https://news.google.com/rss/articles/CBMipAFBVV95cUxQUXZ6MmE5azY5LVZoV0xlM21vWFE2czZIMENGV0h1LTNyazRySExMSmxpNklQTE5BWUU4RlBNWVRTTGhCTnFpVlU5TVBxbmx5Q3E5Yk54X0s5OWJsQ09XeFlZSWtVaTJmNE1mejlXRU9MeHlnbXFPTVBPcTVmUU5aYUdqRElWYUx2TjlMc1RlajB0VGw5UFdMTWs2eHhraUZEcjBsWg?oc=5","source":"search_cn_open_weight_labs","type":"release","summary_1line":"Zhipu identifies Ox Alpha as GLM-5.3-Flash and releases model weights TechNode","sid":"54c9419f0373c186","published":"2026-08-27T07:05:00+00:00","editor_note":"TechNode independently confirms the identity and the weights release under Zhipu's name."},{"title":"GLM-5.3-Flash matches top models at a fraction of the cost, and runs without Nvidia","url":"https://news.google.com/rss/articles/CBMiyAFBVV95cUxQSU5JQmdCR3RyWjhVZUN0aWF6MHhUeEJuaHNvblA3bUNDdl82T1ptY1NsTFVjdUFFN2ZUanZMVEpuY1l3U2ZFQm9zOU1vWU1iVWk2Z0ZjbTlBNUwzeVBSdmhnajRzZm1KemhkOW83ZlJmX3o2NzNzSEZfQWc2YnFiNzM0VDhVOGFWbUJXZU44VDhrNW5JT1RuLTdubUJlMWgydEx2S0g0bEx4TWZfeGRtRWNNQ3N3TlBSZVl0V05BQkZpLWlrdlFMMw?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"GLM-5.3-Flash matches top models at a fraction of the cost, and runs without Nvidia the-decoder.com","sid":"5f95a73de65c4e0a","published":"2026-08-27T10:29:21+00:00","editor_note":"the-decoder.com reports GLM-5.3-Flash matches top models at a fraction of the cost and runs without Nvidia hardware."},{"title":"Z.AI's Use of Chinese Chips for New Model is About Optimization","url":"https://news.google.com/rss/articles/CBMilAFBVV95cUxPRl9KNXFkRnVUdjBKOWhuU3l0VWZ5dkVmNHJoek43eDU0OHBUdU9tNkdMUE9sSThXN2p5YjBEMTUteFhNa0RvamJBWWVjLXVVX3VZUEo2QXNKeWcxXzFGWkZhQUZXaEFacnVCX1Q2X2N2TzVlOWY3bUJFbjk2b2daUWtDNHlBNkN5ak9Wd1o2QkxTblQz?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.AI's Use of Chinese Chips for New Model is About Optimization AI Business","sid":"3cf619be55a479d4","published":"2026-08-27T19:46:33+00:00","editor_note":"A same-day analysis argues the domestic-chip choice is about cost optimization, not a forced workaround."},{"title":"The Powerful Stealth AI Model ‘Ox Alpha’ Is Now GLM-5.3-Flash, and You Can Use It","url":"https://news.google.com/rss/articles/CBMixgFBVV95cUxQQTVuRFM1dUZuS3dma3M3N2dUQ25HYUNlUjJBeWY5c0Rib3V6MkVjVnpJYXJQeDlWZ0tjVkUxLWppRnNxQkZLRkU2MFhpOWJfVXFWazczTmJoNFZUekhPMGxSTlFibjIyMDUwOGhOeG9VOWNabDZjdWJ0Vm1ITTJMOEFhVElhNzZiWWNCTnZnc2hNeFNvemExYi05Z01uN25UU0ZQdVpEZ253RkJkZXQ3VURfZFhPR2lWamlWTlc4Z192WXp0QXc?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"The Powerful Stealth AI Model ‘Ox Alpha’ Is Now GLM-5.3-Flash, and You Can Use It CNET","sid":"0ba8eefddaf4551a","published":"2026-08-27T21:17:10+00:00","editor_note":"CNET shifts the story from identity to availability: Ox Alpha is GLM-5.3-Flash and readers can use it now."},{"title":"Silicon Valley Sensation Ox Alpha Unveiled as Chinese AI Model","url":"https://news.google.com/rss/articles/CBMiiAFBVV95cUxNbHdMWTRZa1FPTW11aGNfeUEyZlVYb0JlWndBRWNIdWw3R0tHOUVKWjE4MWxyRmpNamRKRFlwY3dOVmctckNSVVZvNkNJd0Z1Nk9wVGlhRFRZMmxEM2xMbkFqSW43ZDdhclFibThTbl9CdDBzb3FFQk5EcXlPaU1YVGdiRHdILXAx?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Silicon Valley Sensation Ox Alpha Unveiled as Chinese AI Model 조선일보","sid":"708063be727cbd74","published":"2026-08-27T22:03:37+00:00","editor_note":"Coverage frames the Ox Alpha reveal as a Silicon Valley talking point, not just a regional story."}]},{"date":"2026-08-28","items":[{"title":"Z.ai Unveils Multimodal GLM-5.3-Flash With Lower Serving Costs","url":"https://news.google.com/rss/articles/CBMitgFBVV95cUxPVkt3TlJHNHRMdXRfNEczYjNmVk5YWXVVSTNKZzBwSGtsY014SFhFbWVVUUhuMnlpc1Z4aURjZTMyeE5KLUhJU0ZGZTltYkMxeFpwS2R5Q0JaWVcydW9ubWFkNVN3bHBzd1hyYXlkdmtELTV0RlRLM29LemUzZjhmR3B3UEhJb1h1NDRyZUdaWWlLeXV0NkxzeVR1SFFmWFNSdjRCQm5uR0t3RlB4WjZPSHNidTRldw?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.ai Unveils Multimodal GLM-5.3-Flash With Lower Serving Costs marketscreener.com","sid":"bc0b2cd93677aa22","published":"2026-08-28T02:23:21+00:00","editor_note":"A follow-up highlights GLM-5.3-Flash's multimodal capability and its lower serving costs specifically."},{"title":"Everything we know about Z.ai, the Chinese company behind the mysterious Ox Alpha model","url":"https://news.google.com/rss/articles/CBMikgFBVV95cUxQeF8xa2thZ1NJSC1QWkVPZTRiVFVuS0tpdFdyT3IxRjJabmp4aTh1M1ZCdlRQQzhKWFZCaEYtUGExV3RHcWVieEdHSS04ZHZIUGVXWmMyeVZVcnZQUW5Oclh2aGlLbGYyTW5YYndJRXdIMExsUVZxZG0yVXdvZlNzUFFEemstdGVkd1licGRvTFJhQQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Everything we know about Z.ai, the Chinese company behind the mysterious Ox Alpha model Business Insider","sid":"50a675f74133978a","published":"2026-08-28T07:29:13+00:00","editor_note":"A recap piece rounds up everything known about Z.ai and the Ox Alpha reveal, folding in the domestic-chip detail."},{"title":"Z.ai reveals Ox Alpha was GLM-5.3-Flash, built on Chinese chips","url":"https://news.google.com/rss/articles/CBMibkFVX3lxTE5OdzZMZnU3amFXTWJjb0YxOGVOdnJQdUdrTEI3Y2wxbVN0NGhyYWVzdGpiWThJTmsyUmJzRGpWMGxxSGR4eDM2TEd2N2JjVVp4ZTdzcEV0M0dFMXR0T1NtdDdIWUFZLWNVV3BXY0ZR?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.ai reveals Ox Alpha was GLM-5.3-Flash, built on Chinese chips qz.com","sid":"18b8bf131447c27f","published":"2026-08-28T11:42:10+00:00","editor_note":"A second outlet confirms Ox Alpha's domestic-chip build directly, closing the loop the recap opened."}]},{"date":"2026-08-30","items":[{"title":"China's GLM-5.3 Nears Anthropic on Cyber Test [2026]","url":"https://news.google.com/rss/articles/CBMidkFVX3lxTE1Qc1BxNFFZRmNJdDhsNTU2dFFuZ3dEeHZISURRUERjRHVhOF9nM2ZGb0p2dmNLdW1yTG1OWllvOVh2OUp1SmdxTjJYVEtod0JGdUx2Z0RUMVY0dnUtTThXbVQwZ1pUeHBqN2Rxek5UaG5nWW9fSkE?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"China's GLM-5.3 Nears Anthropic on Cyber Test [2026] tech-insider.org","sid":"400c854a29b0b02a","published":"2026-08-30T00:22:57+00:00","editor_note":"A benchmark write-up puts GLM-5.3's cyber-test score nearing Anthropic's models, without detailing the test methodology."},{"title":"Ox Alpha Emerged as a Powerful OpenAI Rival—Then China's Z.ai Was Revealed as Its Creator","url":"https://news.google.com/rss/articles/CBMirwFBVV95cUxPTjJpNGZKRzJHM2tlanV1ZFBTRzZDT09faVV1b2ZpYUtaRmUySVJHM1hvdHc4Q1RYNU05WXNDVFNkelFVdUdsejdWQXozeFBqRW1GdV9hN09xcEtCY2ZLZ1JjOGpNZFFsdzgtZmtCMkxSZktZeGdoSVRDSmo1QllzamdacFJQOUlIZFBIVkVNT0lCanVIYXNFOTEyX0s1NjJPRFlicHhYU0FSdG0yYVZZ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Ox Alpha Emerged as a Powerful OpenAI Rival—Then China's Z.ai Was Revealed as Its Creator International Business Times","sid":"8ac028cd69ad870a","published":"2026-08-30T11:28:15+00:00","editor_note":"International Business Times recaps the full arc, framing Ox Alpha as beating OpenAI before Z.ai's identity was revealed — no new facts, but signals wider syndication reach."}]},{"date":"2026-08-31","items":[{"title":"Zhipu AI's GLM-5.3-Flash Topped Global AI Calls; China Led Volume for 18th Week","url":"https://news.google.com/rss/articles/CBMid0FVX3lxTE8wZEZhc2NXNENfMlV2ekk5WTdfNW55OGo4NHJDdXZrWWp4VHQzWndaTWZtUkp0MHlsaFhXTFI2QUNDbzQ0T3pBeDFVT3ktTnNkRWxEeVh3dnpQa21vQ3JRSHFZc1p6SUtieHVGTENzd2hIUzJYMlMw?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Zhipu AI's GLM-5.3-Flash Topped Global AI Calls; China Led Volume for 18th Week Pandaily","sid":"262d27de3c444a60","published":"2026-08-31T08:06:18+00:00","editor_note":"Pandaily reports GLM-5.3-Flash topped global AI call volume, with China leading weekly usage for an 18th straight week."},{"title":"Z.AI's Secret AI Model Ran on Chinese Chips — Nobody Noticed","url":"https://news.google.com/rss/articles/CBMiZkFVX3lxTE9oQjlBQVBDZU5jN0hKeXJLbURIZ0hPZVFGOExydnlZRzNwZ3RxcDhhNHhzQ3lYaHF3ODAtb1RqcW0yYkYwUThoZkxXZ3I4c1p3VjdWOGJpbkJLdTBELVRYTEtsczc4UQ?oc=5","source":"search_cn_open_weight_labs","type":"news","summary_1line":"Z.AI's Secret AI Model Ran on Chinese Chips — Nobody Noticed Memeburn","sid":"01ff32093001c421","published":"2026-08-31T10:08:21+00:00","editor_note":"A retrospective says the domestic-chip build behind GLM-5.3-Flash went unnoticed for weeks before the Aug 26-28 reveal made it public."}]}],"editorial":{"tldr":"Zhipu quietly built out a 50,000+ Chinese-chip inference base in mid-August while releasing GLM-5.3 with a claimed 50% coding jump, as a mysterious \"Ox Alpha\" model surfaced in benchmarks rivaling DeepSeek with no confirmed maker. Z.ai then confirmed Ox Alpha was its own model and open-sourced it as GLM-5.3-Flash, a 320B-A18B natively multimodal MoE with a 1M-token context.","stale":false,"whats_new":"Coverage keeps confirming the real story behind the reveal: GLM-5.3-Flash runs on 100% domestic Chinese chips, matches top models at a fraction of the cost without Nvidia hardware, and a retrospective says that chip build went unnoticed for weeks before the reveal made it public.","why_it_matters":"GLM-5.3-Flash's lower serving costs are tied to its domestic-chip inference path, not just open weights — evaluate it on that basis, and don't assume Western-GPU pricing or performance parity without your own benchmark run.","take_for_builders":"Re-run any evaluation you did against the unbranded \"Ox Alpha\" weights using the now-official GLM-5.3-Flash release, and confirm whether its lower serving costs and no-Nvidia claim hold on your own hardware before assuming cost parity.","status":{"state":"Open-sourced · running on domestic chips","tone":"resolved","changed":"2026-08-28","detail":"Ox Alpha shipped as the open-source GLM-5.3-Flash on Aug 26-27, corroborated by four outlets, and Aug 27-28 coverage confirmed it runs on domestic Chinese chips at a fraction of the cost of Western models. Aug 30-31 add a cyber-benchmark claim and a retrospective on the build going unnoticed for weeks.","track":[{"label":"50K+ Chinese chips added","detail":"Aug 12","tone":"rising","weight":6},{"label":"GLM-5.3 launches, coding jumps 50%","detail":"Aug 14","tone":"launch","weight":8},{"label":"early cyber/coding claims","detail":"Aug 18-19","tone":"rising","weight":6},{"label":"Ox Alpha confirmed as Z.ai's own","detail":"Aug 26","tone":"turn","weight":12},{"label":"open-sourced as GLM-5.3-Flash","detail":"Aug 26-27","tone":"rising","weight":18},{"label":"domestic-chip reveal, low cost, stock surge","detail":"Aug 27","tone":"rising","weight":20},{"label":"multimodal follow-up, lower costs","detail":"Aug 28","tone":"now","weight":8},{"label":"cyber-benchmark + unnoticed retrospective","detail":"Aug 30-31","tone":"now","weight":7}]},"beats":[{"kicker":"CHIP FOOTHOLD","tone":"launch","headline":"Zhipu adds 50,000+ Chinese AI chips as its API base nears 7 million users","summary":"Weeks before Ox Alpha has a name, Zhipu is already building the domestic-chip inference capacity that later turns out to be the reveal's real story.","sids":["fde0e6704e2e8662"]},{"kicker":"LAUNCH","tone":"launch","headline":"Zhipu releases GLM-5.3, claiming a 50% coding jump and the strongest open-weights coding model","summary":"Two outlets cover the initial GLM-5.3 release the same day, including a promise that open-source weights would follow within two weeks.","sids":["1359156e75b44e45","842cf30c053df9d2"]},{"kicker":"CYBER CLAIMS","tone":"rising","headline":"Early coverage claims GLM-5.3 already rivals US models on cyber tests, then flags a Cursor bug","summary":"Cyber Magazine reports GLM-5.3 beating US models on cyber-race benchmarks and a second outlet reads the same results past the headline number; a day later Z.ai's own results add coding-test gains but flag a vulnerability in Cursor.","sids":["7216c7465c1c56b5","d209b402e386f93b","6476118b701d8fd0"]},{"kicker":"THE REVEAL","tone":"turn","headline":"Z.ai confirms the mysterious \"Ox Alpha\" stealth model is its own, weights coming","sids":["a9c204e117f936d5"]},{"kicker":"OPEN-SOURCED","tone":"rising","headline":"Ox Alpha ships as GLM-5.3-Flash, a 320B-A18B multimodal MoE with 1M-token context","summary":"The delayed release lands under a different name, and four outlets independently pin the identity: Zhipu names Ox Alpha as GLM-5.3-Flash, publishes the weights, and coverage confirms the model is usable today.","sids":["05d40e9a614e0aad","135778ccf99d98eb","54c9419f0373c186","0ba8eefddaf4551a"]},{"kicker":"THE TURN","tone":"rising","headline":"Ox Alpha's real story: domestic Chinese chips at a fraction of the cost, not just a new name","summary":"Coverage reveals the model runs on 100% domestic chip support, a Chinese-language deep dive prices it near 1/40 of Opus 4.8, Global Times reports the release is powered by 100,000 domestically made chips, and Z.ai-linked shares jump 8%. A same-day analysis calls the chip choice optimization, not necessity; a follow-up confirms it matches top models without Nvidia hardware.","sids":["5bebd783d4b4b151","bc33de06305fc366","7e30319c44f49483","3cf619be55a479d4","708063be727cbd74","5f95a73de65c4e0a"]},{"kicker":"NOW","tone":"now","headline":"Follow-ups, recaps, and a fresh benchmark claim keep the story spreading","summary":"A follow-up highlights GLM-5.3-Flash's multimodal capability at lower serving costs, two recaps round up the full Ox Alpha story, a second outlet confirms the domestic-chip build directly, an Aug 30 benchmark claim puts GLM-5.3 nearing Anthropic on cyber tests, and Aug 31 data shows it topping global AI call volume as China led weekly usage for an 18th straight week.","sids":["bc0b2cd93677aa22","50a675f74133978a","18b8bf131447c27f","8ac028cd69ad870a","400c854a29b0b02a","262d27de3c444a60"]},{"kicker":"STILL UNNOTICED","tone":"now","headline":"A retrospective says the domestic-chip build went unnoticed for weeks before the reveal","sids":["01ff32093001c421"]}],"open_questions":["Does running on Chinese chips carry any export-control or licensing exposure for teams outside China adopting GLM-5.3-Flash?","Do the published weights run unmodified on Western GPUs, or does the release assume a domestic-chip runtime?","What benchmark and methodology backs the Aug 30 claim that GLM-5.3 nears Anthropic on cyber tests?"],"provenance":{"a9c204e117f936d5":{"surfaced_by":"scout"},"05d40e9a614e0aad":{"surfaced_by":"scout"},"135778ccf99d98eb":{"surfaced_by":"scout"},"bc33de06305fc366":{"surfaced_by":"scout"},"7e30319c44f49483":{"surfaced_by":"scout"},"54c9419f0373c186":{"surfaced_by":"scout","verified":4},"5f95a73de65c4e0a":{"surfaced_by":"scout"},"3cf619be55a479d4":{"surfaced_by":"scout"},"0ba8eefddaf4551a":{"surfaced_by":"scout"},"708063be727cbd74":{"surfaced_by":"scout"},"400c854a29b0b02a":{"surfaced_by":"scout"},"262d27de3c444a60":{"surfaced_by":"scout"},"01ff32093001c421":{"surfaced_by":"scout","status_update":true}},"generated_at":"2026-09-02T05:10:00+00:00"}}