Z.ai

Z.ai (Zhipu AI) is the developer of the ChatGLM family of large language models and related AI systems for language, vision, and agent-based tasks.
They focus on building powerful LLMs, multimodal AI, and tools for training, inference, and evaluation.

Conversational AI code generation image generation video generation multimodal reasoning AI agent development long-context processing AI benchmarking autonomous task execution

Tool Information

Primary Task	Chat
Category	communication-and-support
Sub Categories	chat-and-conversation text-to-code text-to-image text-to-video workflow-automation
Pricing	Free and Paid (varies by deployment and model usage)
Country	United States

Z.ai, also known as Zhipu AI, is a leading AI research and development organization specializing in the creation of advanced large language models and AI agents. Their flagship product line includes the ChatGLM series, along with specialized models such as CodeGeeX for code generation, CogVLM (VisualGLM) for visual-language tasks, and CogView/CogVideo for image and video generation. They also innovate in AI agent technology with solutions like CogAgent, AutoWebGLM, and AgentTuning, enabling LLMs to perform autonomous reasoning and interactive tasks. Beyond model creation, Z.ai develops robust evaluation frameworks such as AgentBench, AlignBench, and LongBench, designed to benchmark AI performance in complex reasoning, alignment, and long-context scenarios. Their work serves researchers, developers, and enterprises seeking state-of-the-art AI solutions across language, vision, and multimodal applications.

智谱华章 is a company based out of 北京市海淀区中关村东路1号院6号楼6层603A, 北京, 北京, China.

Pros
Wide range of specialized models strong multimodal capabilities active research in AI evaluation focus on AI agents open releases for developers

Cons
Some models require high compute resources enterprise features may be tied to specific ecosystems rapid development pace can require frequent updates

AI Models

GLM-4.5 Flagship open-source hybrid reasoning and agentic model with MoE architecture.

Jul 2025
Active

GLM-4.5-Air Smaller, faster variant of GLM-4.5 with hybrid reasoning and tool-use capabilities.

Jul 2025
Active

AutoGLM agent (GLM-Z1-Air / GLM-4-Air-0414 powered Free AI agent (AutoGLM Rumination) for tasks like planning, web search; powered by GLM variants.

Mar 2025
Active

GLM-Z1-Air Reasoning variant powering AutoGLM agent; fast and efficient.

Mar 2025
Active

GLM-4-Air-0414 Foundation model variant powering AutoGLM agent.

Mar 2025
Active

GLM-4.0 (speech LLM) Open-source, end-to-end speech large language model adapting tone, emotion, dialect.

Oct 2024
Active

Ying (text-to-video) Text-to-video model generating ~6-second clips from text/image prompts.

Jul 2024
Active

ChatGLM (series) Pre-trained dialogue model series co-released with Tsinghua KEG; open source.

Nov 2023
Active