---
title: "ZeroShell (@zeroshell) | Hashnode"
description: "ZeroShell (@zeroshell) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@zeroshell"
---

@zeroshell

# ZeroShell

@zeroshell · Joined November 2026

### [《怎么判断一个 API 中转站靠不靠谱：四个维度，和一个真实样本》](https://zeroshell.hashnode.dev/api?utm_source=hashnode&utm_medium=feed)

先说利益相关：下面提到的 ZeroShell 是我自己运营的站，所以这篇重点先讲怎么判断一个中转站靠不靠谱，结论你自己下。 判断中转站，我看四点： 第一，模型目录是不是按分组算的。 很多站宣传「几十个模型」，但那是全站聚合数字，单个 Key 能调的只是它所在分组的目录。建错分组，用 Claude 时直接报 model\_not\_found，新手最容易栽在这。 第二，价格是否公开明码。 敢把每个模型每

1 min read · [zeroshell.hashnode.dev](https://zeroshell.hashnode.dev/api?utm_source=hashnode&utm_medium=feed)

[Nov 4](https://hashnode.com/posts/api/6aebc20b55cc06a78ec2a307)

### [Run Claude Code on a Relay Endpoint: Setup, the Model-Catalog Pitfall, and What It Actually Costs](https://zeroshell.hashnode.dev/run-claude-code-on-a-relay-endpoint-setup-the-model-catalog-pitfall-and-what-it-actually-costs?utm_source=hashnode&utm_medium=feed)

Most AI coding tools speak one of two dialects: the OpenAI chat-completions format, or Anthropic's Messages format. If you want to run those tools against a third-party relay endpoint instead of a fir

5 min read · [zeroshell.hashnode.dev](https://zeroshell.hashnode.dev/run-claude-code-on-a-relay-endpoint-setup-the-model-catalog-pitfall-and-what-it-actually-costs?utm_source=hashnode&utm_medium=feed)

[Nov 4](https://hashnode.com/posts/run-claude-code-on-a-relay-endpoint-setup-the-model-catalog-pitfall-and-what-it-actually-costs/6aeb559915deb0a68d9167cc)