←
Matthew Young8 posts
Matthew Young avatar

Matthew Young

@OpenRouterMatt
◷ Joined June 2025
Chris Clark avatar
⟳ Matthew Young retweeted
Chris Clark@cclark·Nov 5

Re: #3 -- Wait, WHAT?!? All we do is complain about the Azure portal UI. Internally, externally, to MSFT teams. Every company I interact with who uses Azure talks about the terribleness of the portal. It's a running joke *with our support team at Microsoft*. We have long support calls where all we are trying to do is navigate the portal. We have literally to moved things onto other clouds because no one can figure out the how subscriptions/regions/resource groups/projects etc interact. If somehow someone influential is reading this, I *beg* you to *please* improve the portal!

OpenRouter avatar
⟳ Matthew Young retweeted
OpenRouter@OpenRouter·Nov 3

In a year, AI on OpenRouter went from a few people per company to entire departments. People using it on teams of 10+ grew 70x, and keys per org tripled. That's a lot more to manage. So we've updated Guardrails, the controls admins use to set AI budgets, model access, and content moderation across their company. Now you can set restrictions for the whole org, give specific people bigger budgets or more models within those limits, and change each policy without touching the others.

Post attachment
OpenRouter avatar
⟳ Matthew Young retweeted
OpenRouter@OpenRouter·Jun 15

New Free capacity on OpenRouter, served by @eigenlabs's Darkbloom: gpt-oss-20b and Gemma 4 26B Start using the models today ↓

Post attachment
Gajesh avatar
⟳ Matthew Young retweeted
Gajesh@gajesh·Jun 11

Monday: Darkbloom goes live on OpenRouter. Every OpenRouter user gets a free trial of private inference, powered by idle Macs around the world. And for providers -- next week we're launching an Alpha Rewards Program: earn up to $40 guaranteed just for running a node. Limited time while we're in alpha. Breakdown in the replies. The network is ready for it. Since our Public Alpha two weeks ago: Gemma 4 is now multimodal, SSD caching has made time-to-first-token much faster, and load tests show we can sustain millions of tokens per second -- with more upgrades landing before launch. If you run a node: turn it on. Demand arrives Monday. If you have a Mac: darkbloom(.)dev → one install command and you're earning. The world's sleeping compute is waking up.

2:04
Infisical avatar
⟳ Matthew Young retweeted
Infisical@infisical·Apr 10

How do you give Cursor, Claude Code, and Devin access to secrets without handing them the keys to production? @OpenRouter had to figure this out. 70+ model providers. New ones added every week. Each one bringing API keys that needed to reach specific services. For a while, custom scripts and manual updates held it together. Then their engineers started shipping with AI agents, and the old system couldn't answer a basic question: which credentials should an agent get, and which should it never touch? They built three access tiers with Infisical: → Dev secrets for engineers → Session-scoped credentials for AI agents → Locked-down access for production workloads Adding a new provider key went from a multi-service chore to a single update. 50+ engineering hours saved per month. Full story from OpenRouter's founding engineer: infisical.com/customers/open…

OpenRouter avatar
⟳ Matthew Young retweeted
OpenRouter@OpenRouter·Apr 1

We asked ourselves: what if the best UI era never ended? Introducing OpenRouter's new look, inspired by the golden age of the web 🌐 ✨ Under construction. Always. 🚧

Post attachment