zai-org/glm-4

GLM-4 series: Open Multilingual Multimodal Chat LMs | 开源多语言多模态对话模型

View on GitHub ↗Jump to charts ↓

Summary Information

Updated 58 minutes ago
Added to GitGenius on September 9th, 2026
Created on May 15th, 2024
Open Issues & Pull Requests: 35 (+0)
GitHub issues: Enabled
Number of forks: 612
Total Stargazers: 7,072 (+0)
Total Subscribers: 48 (+0)

Repository Insights (GitGenius)

Most active contributors

Sign in to see contributor activity.

Related repositories by overlapping contributors

No overlapping-contributor repos identified yet.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Open issues: 35
New in 7 days: 0
Closed in 7 days: 0
Avg open age: 391 days
Stale 30+ days: 34
Stale 90+ days: 33

Recent activity

Opened in 7 days: 0
Closed in 7 days: 0
Comments in 7 days: 0
Events in 7 days: 0

Top labels

  • documentation (1)
  • help wanted (1)

Most active issues this week

No issue events were indexed in the last 7 days.

Detailed Description

GLM-4 is an open-source multilingual multimodal chat language model series that supports local deployment and serves as an alternative to proprietary large language models.

The project addresses the need for capable open-source language models by releasing the GLM-4-32B-0414 series, which includes base models and specialized variants. The base model was pre-trained on 15 trillion tokens of high-quality data, including reasoning-type synthetic data, then refined through human preference alignment and reinforcement learning techniques. This approach targets improvements in instruction following, code generation, function calling, and complex reasoning tasks. A specialized reasoning variant, GLM-Z1-32B-0414, adds deep thinking capabilities through extended reinforcement learning focused on mathematics, code, and logic problems.

Developers should consider this project if they need an open-source model comparable to GPT-4o or DeepSeek-V3 that can run locally without proprietary service dependencies. The tool suits applications requiring strong performance in code generation, function calling for agent tasks, artifact generation, and complex reasoning. The project explicitly positions itself against commercial offerings by providing free access to open-source models through a hosted interface while maintaining the ability for local deployment.

The project maintains active development with regular model releases across different capability tiers and parameter scales. The team publishes technical reports documenting their approach and releases multiple model variants addressing different use cases, from base models to specialized reasoning versions. Community engagement is supported through multiple channels including Discord and social media platforms.