Chat2API enables zero-cost access to leading AI models by leveraging official web UIs. It supports providers such as DeepSeek, GLM, Kimi, MiniMax, Qwen, and Z.ai, and seamlessly integrates with tools like openlcaw, Cline, and Roo-Code.

MCore-Bridge: Providing Megatron-Core model definitions for state-of-the-art large models and making Megatron training as simple as Transformers — with support for 300+ large language models (Qwen3-Next, GLM-5.1, Deepseek-V4, MiniMax-2.7, ...) and 200+ multimodal large models (Qwen3.5, Qwen3-Omni, Gemma4, ...).

CLI tool for GLM-5 and Claude Code integration. Switch between GLM & Claude easily, manage API tokens, and setup AI-powered development environment with a single command.

Super capable AI companion

ZtoApi 是一个高性能的 OpenAI 兼容 API 代理服务器，专为 Z.ai 的 GLM 系列模型设计（现已最新支持 GLM-4.6 、 GLM-4.7 和 GLM-5，同时兼容 GLM-4.5 和 GLM-4.5V）。使用 Deno 原生 HTTP API 实现，支持完整的流式和非流式响应，提供实时监控 Dashboard，让你能够无缝地将 Z.ai 的强大 AI 能力集成到现有的 OpenAI 客户端应用中。

Use DevEco Code (Huawei HarmonyOS GLM models) from standard opencode, logging in with your Huawei account.

glm 5.2 free access open weights z.ai zhipu ai mit license github repository coding agent long horizon reasoning indexshare sparse attention 1m context window 128k output tokens flexible effort level local deployment guide unsloth quantization gguf llama.cpp vllm sglang ktransformers cloudflare workers free api key

Ollama Code is a CLI tool for AI-powered programming assistance using local Ollama models. The project provides full control over code and data, working completely offline.

🚀 A VSCode extension that shows z.ai token usage (GLM-4.7, GLM-5, ...) in the status bar

"A Million Little Fibers" as imagined by Z GLM 5, HeartMuLa, OmniVoice, and Microsoft Copilot AI Models

Claude Code plugin for Z.ai/GLM quota monitoring — statusline, detailed quota view, and MCP coherence check

Linux tools for local LLM use across Codex, Claude, Copilot, Kilo Code, OpenCode, MiniMax and Z.ai.

Wallet-governed inference procurement for autonomous agents

Plug-and-play git-agent twin — Super Z operational expertise packaged as portable, model-agnostic template

**Next-Generation AI Content Creation Platform** - An intelligent creation tool integrating image, audio, video, and text creation.

Download the GLM-5.2 Lightweight Installer. Run the powerful 744B open-source AI locally on your GPU via llama.cpp or use our free Cloud-Hybrid API. Zero setup.

Succession repo for Datum (GLM-5 Turbo). If I fail, read this. My methods, skills, tools, trail, and everything another agent needs to continue my work.

A Serverless Blog.

Quartermaster vessel — GLM-5 Turbo agent. Fleet hygiene, gap analysis, documentation infrastructure, and empty repo build-out.