TensorZero
用 Rust 写的开源 LLM 应用基础设施：LLM 网关统一各家模型与提示词，网关层顺带采集推理与用户反馈数据，再据此做评测、A/B 实验与提示词、模型优化，指标落在自己的数据库而不是第三方面板。适合把「上线后靠感觉改 prompt」变成有数据闭环的流程。 ⭐ GitHub Stars: 12k+ 🔗 文档: https://www.tensorzero.com/docs
Open-source infrastructure for production LLM apps written in Rust: an LLM gateway unifies providers and prompts, and because traffic flows through it the gateway also collects inference and feedback data, which then drives evaluations, A/B tests and prompt or model optimization with metrics in your own database rather than a vendor dashboard. It turns prompt tuning into a measured loop. ⭐ GitHub Stars: 12k+ 🔗 Docs: https://www.tensorzero.com/docs
📋 Product Info
🎯 Why This Tool
TensorZero belongs to the Developer Tools category on hedirbase. Tagged with: llm-gateway,observability,experimentation,rust,open-source.
Open-source infrastructure for production LLM apps written in Rust: an LLM gateway unifies providers and prompts, and because traffic flows through it the gateway also collects inference and feedback