LangChain 做了一款给团队管住编码 AI 花钱的东西，能对接 Cursor、Codex 和 Claude Code，还能设预算上限，谁用谁知道。
LangChain 创始人 Harrison Chase 宣布其 LLM Gateway 进入私有预览，重点解决编码智能体（如 Cursor、Codex、Claude Code）带来的成本激增问题。该网关提供准确的模型定价（考虑缓存、令牌层级等变量），集成多种开发工具，并允许设置成本上限与动态调整。团队反馈显示，6个月内编码智能体支出增长显著，该方案旨在让企业同时赋予开发者选择自由和财务可控性。
Like many AI-pilled orgs, our spending on coding agents has increased pretty dramatically over the p...
Like many AI-pilled orgs, our spending on coding agents has increased pretty dramatically over the past few months A key part of the LLM gateway we're building (available in private preview now) is cost control and visibility Here's how we're doing it: 1/ Accurate, up to date model pricing. Model pricing is suprisingly non-trivial, with caching, token tier, and other variables coming into play 2/ A lot of integrations. We've had to figure out how to integrate with Cursor, Codex, Claude Code and more (we give our team a lot of choice). We want to continue to let engineers choose which coding agents they use while also managing their spend 3/ Cost controls. Visibility isn't enough, we want a want to control costs. But also allow for bumping any limits if it turns out developers are using it in a proper way We think this type of cost control is something every company (startup or enterprise) will want. If you want to try our the private preview of the LLM gateway that enables this, sign up here: langchain.com/langsmith-llm-… 💬 6 🔄 2 ❤️ 18 👀 1910 📊 9 ⚡