AI Agents
coze-studio
A visual all-in-one platform for building, debugging, and deploying AI agents.
★ 21.7k starsApache-2.0
View guide →A full-lifecycle platform for developing, debugging, evaluating and monitoring AI agents.
coze-loop is an open-source ai agents project by coze-dev (coze-dev/coze-loop on GitHub). It has 5,765 GitHub stars and 801 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 8 October 2026.
Identify what you need to improve — prompt quality, evaluation coverage, or visibility into failed runs.
Clone the repo and prepare the Docker Compose deployment under release/deployment.
Edit model_config.yaml with your model provider's API key and endpoint/model ID.
Bring the services up and open the web interface to develop prompts, run evaluations and inspect traces.
Write, debug, optimize and version prompts in a visual playground with real-time interactive testing and side-by-side comparison.
Automated multi-dimensional evaluation of prompts and agent outputs — accuracy, conciseness, compliance and more.
Records the full execution process from user input to AI output, including prompt parsing, model invocation and tool execution, capturing intermediate results.
Development, debugging, evaluation and monitoring live together instead of in separate tools.
Runs via Docker Compose for a quick local deployment or via a published Helm chart on Kubernetes.
LLM integration is provided by the CloudWeGo Eino framework team, with a Go backend.
Clone Coze Loop and enter the project directory.
git clone https://github.com/coze-dev/coze-loop.git
cd coze-loopEdit the Docker Compose model config with your provider's API key and model/endpoint ID.
# Edit release/deployment/docker-compose/conf/model_config.yaml — set api_key and modelBring Coze Loop up with the provided make target and open the UI.
make compose-up
# Then open http://localhost:8082Coze Loop is maintained by coze-dev, the team behind Coze, with the Eino framework providing LLM integration.
The open-source version is Apache-2.0 licensed and free to self-host; Coze also offers hosted products separately.
Yes — you configure a model provider (the examples use Volcengine Ark / BytePlus ModelArk) with your own API key in model_config.yaml.
Those focus mainly on tracing and evaluation; Coze Loop covers the full lifecycle including a prompt development playground and version management, in a self-hostable Go platform.
Clone the repo, set your model in model_config.yaml, run make compose-up, and open http://localhost:8082.
A visual all-in-one platform for building, debugging, and deploying AI agents.
A framework for evaluating and improving AI agents.
Cross-platform desktop panel for managing OpenClaw and Hermes AI agents.
Management console for installing and monitoring AI agents and pipelines.
SimWorld: an Unreal Engine 5 simulator for evaluating LLM/VLM agents in physical and social worlds.
Build, run, and manage full agent platforms — not just single agents.
The most-starred other AI Agents projects in the LibreAI catalogue — the realistic shortlist if coze-loop is not the right fit.
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
Free and open — grab the source, star the repo, and start creating.