Portkey

Hermes Agent by Nous Research now works with Portkey as a drop-in gateway. Hermes runs on your server across CLI, Telegram, Discord, Slack, and WhatsApp. When it connects to MCP servers, each session manages its own credentials for every server it talks to. Once the agent has access, it has access to everything that server exposes. Portkey sits between Hermes and your MCP servers. Hermes authenticates once. The gateway handles credential injection, decides which tools are visible to the agent, and logs every tool call with the user, parameters, and full response. Your Hermes config doesn't change. You also get access to 3,500+ models, full request logging, cost tracking per session, and budget limits that auto-cut before a session spirals. If you're running Hermes in production, this is worth a look. Docs in comments 

Portkey CLI is live 🚀 One command to wire claude code and codex to your gateway. MCPs configured, skills synced, every developer in your org set up in 60 seconds. npx portkey setup. Docs in comments below!

It’s 2026 and suddenly agents are everywhere. Slowly but surely, AI is moving from pilot to production. The models are ready, but now comes the hard part. As agentic demos move into enterprise deployments, reliability, cost, and observability are taking center stage. In our latest piece, Ethan Volk explores why getting AI to work at scale within the enterprise is a function of the infrastructure and systems that support it. If you’re building the infrastructure that makes AI production ready, we’d love to talk to you. Read the full piece 👉 https://lnkd.in/eRMWsjaj

The Agent Gateway webinar is just a few hours away and registration is still open. We'll show you exactly how Agent Gateway works, live. The architecture, access control scoped per agent and team, policy enforcement without touching your code, and a look at the beta. If you're building with agents, register now : https://lnkd.in/dhWNxnEH

Agent Gateway is in beta. It's the production layer between your agents and the world. Register any agent, get a governed endpoint. Full traces across every run, access control scoped per agent and team, policy enforcement without touching your agent code. Live technical walkthrough this Friday. 📅 April 24 🔗 https://lnkd.in/dhWNxnEH Register now!

The first agent in an org always works. Well-scoped, clear ownership, everyone paying attention. The problems begin somewhere between agent five and agent fifty. Teams lose track of what is running. Permissions drift. Chains fail silently, and by the time anyone notices, the consequences are hard to trace back to the original problem. This is not a failure of planning - It is a governance gap. And it is the most common wall teams hit when scaling agents in production. Our Agent Gateway was built to close this gap. Join us this Friday, April 24, to see how it works. Register: https://lnkd.in/dhWNxnEH

The first agent in an org always works. Well-scoped, clear ownership, everyone paying attention. The problems begin somewhere between agent five and agent fifty. Teams lose track of what is running. Permissions drift. Chains fail silently, and by the time anyone notices, the consequences are hard to trace back to the original problem. This is not a failure of planning - It is a governance gap. And it is the most common wall teams hit when scaling agents in production. Our Agent Gateway was built to close this gap. Join us this Friday, April 24, to see how it works. Register: https://lnkd.in/dhWNxnEH

Portkey's Agent Gateway is now live in beta Register any agent, get a governed Portkey endpoint, and every action flows through a single control plane. Cost controls, observability across 40+ metrics, access control, and 50+ security guardrails — built in from day one. On April 24 we're running a live demo. Register here : https://lnkd.in/dhWNxnEH Docs in comments. See you on Friday! 👋