model
智谱发布 GLM-5.1 高速版 API，刷新全球大模型 API 速度纪录
High-speed but closed Chinese LLM API with impressive performance but limited ecosystem
6.3Overall
Utility7
Onboarding5
Craft6
Niche fit7
Longevity6
Five dimensions scored independently; overall is a weighted average. Scores are only comparable within this same rubric.
Good for
Enterprise clients needing low-latency APIs, especially for AI programming and gaming
Not for
Developers requiring open models or broad ecosystem support
Project description
智谱近日推出 GLM-5.1 高速版 API，模型输出速度达 400 tokens/s，打破高速模型多为轻量级的行业惯例，首次在国产大模型中实现旗舰级能力与低延迟并存。实测显示其在 AI 编程、3D 游戏、交互界面等场景表现出色，核心在于 TileRT 高性能推理引擎，该引擎由智谱 GLM 团队与 TileRT 团队联合打造，通过系统级优化实现。目前适用于速度敏感场景，并通过智谱 MaaS 平台向部分企业客户开放。
Alternatives
GPT-4 TurboClaude 3 OpusDeepSeek-V3