ModelsSelected

Zhipu GLM-5.3-FlashX model released

智谱 GLM-5.3-FlashX 模型上线,更快、更流畅

Why we picked it

For users needing a fast, high-performance AI model, the GLM-5.3-FlashX offers a competitive alternative to other large language models.

Zhipu announced the release of the GLM-5.3-FlashX model today, offering up to 200 tokens per second. The model, previously known as Ox Alpha, is now available via API with a Model Key. It is built on 100,000 domestic chips and aims to provide faster and smoother performance for enterprises and developers.

Original · IT之家

智谱 GLM-5.3-FlashX 模型上线,更快、更流畅

IT之家 9 月 18 日消息,智谱今日宣布推出 GLM-5.3-FlashX(最高 200 tokens/s) ,为企业与开发者带来更快、更流畅的模型体验。 智谱官方表示,GLM-5.3-Flash 此前以“Ox Alpha”之名与全球开发者见面,获得海内外开发者的广泛认可,调用量持续攀升。面对快速增长的需求,智谱在 10 万张国产芯片提供的推理算力基础上, 进一步加大对 Infra 侧的投入与推理优化 。GLM-5.3-Flash 长期保持同尺寸最强智能水平,本次提速进一步形成智能、价格、速度的全面竞争力。 GLM-5.3-FlashX API 现已上线,Model Key: GLM-5.3-FlashX 。 IT之家附 API 调用方法: https://docs.bigmodel.cn/api-reference/ 模型-api/对话补全