Z.ai
New
Access GLM-5.2 (1M token context, MIT-licensed) plus CogVideoX and CogView3 from one API — frontier-class language, video, and image generation at highly competitive prices.
Models
★ 4.4(2,600 reviews)freemiumOverview
Z.ai is Zhipu AI's international platform for GLM-5.2 — a 1 million token context model released June 2026 with MIT open-source weights and strong multimodal reasoning across Chinese and English. The API also powers CogVideoX (text-to-video) and CogView3 (image generation), making it one of the few platforms where a single API key covers language, vision, and video. MIT licensing lets teams self-host with no usage fees; the hosted API remains among the most competitively priced frontier models available.
Key Features
- GLM-4 with 128K context window for long-document processing
- Multimodal input — text, image, and video understanding in one model
- CogVideoX for high-quality text-to-video generation
- CogView3 for photorealistic text-to-image
- GLM-4-Voice for natural speech synthesis and understanding
- Competitive API pricing well below comparable Western frontier models
Pros
- • Single API key for language, image, video, and voice — rare in the category
- • GLM-4 handles Chinese-language tasks at a level no Western model matches
- • Among the cheapest frontier-model pricing per token
Cons
- • English reasoning trails GPT-4o and Claude 3.5 on complex tasks
- • Smaller English-language community than OpenAI or Anthropic
- • Documentation and support are primarily in Chinese
Advertisement