Some of this data exists already, but it may not be fully accurate. I will gradually improve and expand it as I continue developing my projects.

Also, if you look at the JSON data, there is some other information as well (not yet shown in the UI), such as what reasoning effort levels different models support, or which providers offer the same protocol, and so on.

Another reason is that there are now more and more Conditional Pricing rules, but various model libraries do not include them, which makes it very difficult to implement pricing properly. So I decided to build one myself.

The main data sources are models.dev and LobeHub, plus manual corrections. The original reason I built this was that the flat list on the models.dev website was really unpleasant to browse.

Just know @useplunk Pay as you go without a minimal charge is incredible

I just think, you need to use AWS (or other EC2 services) only if you need HA (or upgrade but can’t bear downtime). Durable (or backup) is easy, and is pretty cheap on a VPS. But HA is really expensive if you want to achieve without managed services.

I love @DrizzleORM , but I’m using rust Considering some way to bring the DX to rust now

Though I've been using JS for backend development increasingly over the past few years (thanks to Cloudflare Workers), I still only choose it for personal use or Proof of Concept. It's the hardest to scale a JS backend. Its single-thread model almost certainly precludes its use in any enterprise-level scenarios. I also won't use it in any heavy-backend projects. JS is enjoyable to write, and its ecosystem is well-developed. However, once the project succeeds, its only fate is to be completely refactored. (Well, many people won't care too much about this -- If it comes to the point where it needs to be refactored, then there will be enough money to do so. I agree with this, but this Tweet is from a technical perspective, so we won't discuss the business pros and cons.) But why is JS backend so popular now? I think a major reason is Serverless. The scalability issue of JS is almost nonexistent in serverless environments. With edge computing becoming more and more mature(even though Vercel no longer recommends it), using JS for backend tasks actually provides a scalability advantage over other traditional server-based backends. What about the disadvantages of Serverless? I think it's the cost and the latency to connect to other infrastructure, such as databases and caches. Serverless can be divided into two specs: one that uses existing cloud services (such as Cloudflare Workers, Vercel, and AWS Lambda), and another that is internally built using systems like Kubernetes or Nomad. Although these two types are quite different, they only result in three combinations: low cost + high latency (such as Cloudflare Workers), high cost + low latency (such as AWS Lambda and self-hosted solutions), and high cost + high latency (such as Vercel). There is no solution that offers both low cost and low latency. I will be starting a new project later this year that is very backend-intensive. After much consideration, Java might still be the best solution.