GLM-5.1 high-speed version released: 400 tokens/s, flagship capability at the world's fastest API speed
Original title:GLM-5.1高速版:400 tokens/s,顶尖模型跑出最快速度
AI summary
Zhipu launched the GLM-5.1 high-speed API "GLM-5.1-highspeed" for select enterprise customers, with an output speed of 400 tokens/s, setting a new API speed ceiling among global large-model vendors. This version is the first among domestic models to combine flagship capability with extremely low latency. Built jointly by the GLM team and the TileRT team, it uses system-level optimizations including AOT static orchestration, tile-level micro-tasks and 8-GPU NVL topology specialization to ensure 400 TPS is a stable, production-grade capability.
Key Takeaways
- Zhipu launched the GLM-5.
- 1 high-speed API "GLM-5.
- 1-highspeed" for select enterprise customers,with an output speed of 400 tokens/s,setting a new…
- This version is the first among domestic models to combine flagship capability with extremely low…
- Built jointly by the GLM team and the TileRT team,it uses system-level optimizations including AOT…
Quick Facts
- Published
- Updated
- Source
- 公众号:智谱(GLM)
- Category
- Model Releases
- AI score
- 64
在旗舰模型上实现400 tokens/s,打破了高速必为轻量模型的惯例,对编码智能体等实时交互场景提供了不牺牲能力的低延迟方案。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: GLM-5.1 high-speed version released: 400 tokens/s, flagship capability at the world's fastest API speed. May 22, 2026. https://new.nat6.net/en/news/cmsdu107700tzrofzz8996w79/