LLMstudio

TensorOpsAIMPL-2.0活跃维护

把 LLM 应用从原型推进到生产环境的框架，集成了 LangChain、MLflow、OpenAI 和 Vertex AI，方便管理提示词与模型部署。

Framework to bring LLM applications to production

暂无版本390star43fork同步于 2026-11-07 07:29 UTC

LLM应用部署提示词工程 LLMOps 模型服务 MLflow

对比 GitHub 官网 ai langchain llm llmops ml

怎么上手

由 README 提炼，执行前请确认

原文

pip install 'llmstudio[proxy,tracker]'
llmstudio server --proxy --tracker

硬件要求按 README 说明

README 未说明显存或 GPU 要求；本地模型需通过 Ollama 运行，具体硬件取决于所选模型

下载：源码 zip·GitHub Releases（安装包）本站不镜像安装包

项目说明

由 README 自动翻译提炼

没人工点评？写下你的经验

把 LLM 应用推向生产的框架，统一代理多家模型并管理提示词

把 LLM 应用从原型推进到生产环境的框架，集成了 LangChain、MLflow、OpenAI 和 Vertex AI，方便管理提示词与模型部署。

核心能力

通过代理统一接入 OpenAI、Anthropic、Google 的 LLM
借助 Ollama 支持自定义和本地开源模型
提供 Prompt Playground 界面调试与优化提示词
内置监控与日志记录每次请求的用量和性能
支持批量调用、智能路由与故障回退保障可用性

容易踩的坑

需在运行目录自建 .env 并填入 OPENAI_API_KEY、ANTHROPIC_API_KEY、VERTEXAI_KEY 等密钥；proxy 与 tracker 默认分别占用 50001 和 50002 端口；轻量版 pip install llmstudio 不含 proxy 和 tracker 功能

内容由模型根据 README 生成，可能不完整或不准确，以官方文档为准。看原文 README

用户点评

登录用户的真实使用记录，可按版本和硬件筛选

共 0 条

还没有人点评。第一个说出真实体验的人最有价值。

README 原文

未翻译

在 GitHub 打开

LLMstudio by TensorOps

Prompt Engineering at your fingertips

LLMstudio logo

🌟 Features

LLMstudio UI

LLM Proxy Access: Seamless access to all the latest LLMs by OpenAI, Anthropic, Google.
Custom and Local LLM Support: Use custom or local open-source LLMs through Ollama.
Prompt Playground UI: A user-friendly interface for engineering and fine-tuning your prompts.
Python SDK: Easily integrate LLMstudio into your existing workflows.
Monitoring and Logging: Keep track of your usage and performance for all requests.
LangChain Integration: LLMstudio integrates with your already existing LangChain projects.
Batch Calling: Send multiple requests at once for improved efficiency.
Smart Routing and Fallback: Ensure 24/7 availability by routing your requests to trusted LLMs.
Type Casting (soon): Convert data types as needed for your specific use case.

🚀 Quickstart

Don't forget to check out https://docs.llmstudio.ai page.

Installation

Install the latest version of LLMstudio using pip. We suggest that you create and activate a new environment using conda

For full version:

pip install 'llmstudio[proxy,tracker]'

For lightweight (core) version:

pip install llmstudio

Create a .env file at the same path you'll run LLMstudio

OPENAI_API_KEY="sk-api_key"
ANTHROPIC_API_KEY="sk-api_key"
VERTEXAI_KEY="sk-api-key"

Now you should be able to run LLMstudio using the following command.

llmstudio server --proxy --tracker

When the --proxy flag is set, you'll be able to access the Swagger at http://0.0.0.0:50001/docs (default port)

When the --tracker flag is set, you'll be able to access the Swagger at http://0.0.0.0:50002/docs (default port)

📖 Documentation

Visit our docs to learn how the SDK works (coming soon)
Checkout our notebook examples to follow along with interactive tutorials

👨‍💻 Contributing

Head on to our Contribution Guide to see how you can help LLMstudio.
Join our Discord to talk with other LLMstudio enthusiasts.

Training

Thank you for choosing LLMstudio. Your journey to perfecting AI interactions starts here.