{"rewrite":{"id":"r_aa49320da8381d697886b1c0","clusterId":"c_7ae9dda19f032db46d7d16a8","slug":"baidu-releases-ernie-5-5-flash-base-as-an-open-model","model":"deepseek-v4-1-flash","headline":"Baidu Releases ERNIE 5.5 Flash Base as an Open Model","summary":"Baidu released ERNIE 5.5 Flash Base as an open model. It is a mixture-of-experts model with 280 billion total parameters and 14 billion active parameters, pretrained on 33.5 trillion tokens, with a 1 million token context window for inference and fine-tuning. Baidu says it is built so training and execution costs stay low, and that long-context understanding is built in at the foundation stage. A version with additional training, ERNIE 5.5 Flash, is planned for release soon.","whyItMatters":"Baidu is betting that cost-controlled foundation models with long-context handling baked in at the base stage will be easier to extend than models patched after the fact, and it is shipping the base weights openly before the tuned version arrives.","webCardHtml":"\u003cp\u003eBaidu published the weights on Hugging Face under the baidu/ERNIE-5.5-Flash-Base repository. The MoE design keeps 14 billion of its 280 billion parameters active per pass, which is the part that matters for execution cost.\u003c/p\u003e\u003cp\u003eBaidu frames the design around keeping additional training and inference cheap, and around adding abilities like long-context understanding at the foundation stage rather than later. Benchmark results in the release compare it against DeepSeek-V4-Flash-Base, Hy3-preview-Base, and MiMo-V2.5-Base, with ERNIE 5.5 Flash Base scoring above foundation models of comparable scale. A further-trained ERNIE 5.5 Flash is planned for release in the near future.\u003c/p\u003e","blueskyPost":"ERNIE 5.5 Flash Base activates only 14 billion of its 280 billion parameters, so Baidu is selling cost efficiency as the model's headline feature, not raw scale. That choice targets fine-tuning budgets more than benchmark scores.","twitterPost":"ERNIE 5.5 Flash Base activates 14 billion of 280 billion parameters. Baidu is selling cheap training and inference, not raw scale.","threadsPost":"ERNIE 5.5 Flash Base carries 280 billion total parameters but activates only 14 billion per pass, and Baidu pairs that with a 1 million token context window. The pitch is cost, not scale: a base model tuned for cheap fine-tuning ahead of the fuller ERNIE 5.5 Flash.","newsletterBlurb":"Baidu has released ERNIE 5.5 Flash Base as an open model, a mixture-of-experts design with 280 billion total parameters and 14 billion active, pretrained on 33.5 trillion tokens. Baidu says long-context understanding is built in at the foundation stage rather than added later. A further-trained ERNIE 5.5 Flash is planned for release in the near future.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20261106-ernie-5-5-flash-base/\",\"title\":\"China's Baidu Releases High-Performance AI Model \\\"ERNIE 5.5 Flash Base,\\\" a Foundation Model Designed for Future Enhanced Versions\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":5621,"outputTokens":856,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1793955596,"createdAt":"2026-11-06T08:54:52.000Z","publishedAt":"2026-11-06T08:58:51.000Z","updatedAt":"2026-11-06T08:58:51.000Z"},"cluster":{"id":"c_7ae9dda19f032db46d7d16a8","canonicalTitle":"中国のBaiduが高性能AIモデル「ERNIE 5.5 Flash Base」を公開、今後の強化版を見据えた基盤モデル","representativeArticleId":"a_80f98ec239577e4bcd603053","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[\"ERNIE 5.5 Flash Base\"],\"studios\":[\"Baidu\"],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-11-06T08:12:00.000Z","lastSeenAt":"2026-11-06T08:12:00.000Z","updatedAt":"2026-11-06T08:58:53.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20261106-ernie-5-5-flash-base/","title":"中国のBaiduが高性能AIモデル「ERNIE 5.5 Flash Base」を公開、今後の強化版を見据えた基盤モデル"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":["ERNIE 5.5 Flash Base"],"studios":["Baidu"],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":["Baidu released ERNIE 5.5 Flash Base as an open model.","ERNIE 5.5 Flash Base is a MoE model with 280 billion total parameters and 14 billion active parameters.","ERNIE 5.5 Flash Base was pretrained on 33.5 trillion tokens and supports a 1 million token context window.","Baidu plans to release a further-trained ERNIE 5.5 Flash in the near future.","The release compares ERNIE 5.5 Flash Base against DeepSeek-V4-Flash-Base, Hy3-preview-Base, and MiMo-V2.5-Base."],"pageLinks":{"healthSlugs":["ernie-5-5-flash-base","baidu"],"storySlugs":[],"entities":[{"slug":"ernie-5-5-flash-base","supported":0,"disputed":0,"contradicted":0,"unverified":3,"overall":"unverified","story_count":1,"has_dossier":false},{"slug":"baidu","supported":0,"disputed":0,"contradicted":0,"unverified":1,"overall":"unverified","story_count":1,"has_dossier":false}],"stories":[],"entitiesOk":true,"storiesOk":true}}