Skip to content
CC Switch WikiUnofficial Community Wiki

x-anthropic-billing-header breaks caching - CC Switch

Quick fix

Upgrade CC Switch to v3.15.0+, which fixes the dynamic billing header breaking prefix caching in openai_responses mode.

Symptom

Section titled “Symptom”
Error output
上次自动同步失败：Fetch is aborted
English translation

上次自动同步失败：Fetch is aborted

Cause

Section titled “Cause”

This cluster actually mixes several quite different CC Switch request conversion and proxy defects. This page addresses the fixed core problem (#2350): in `openai_responses` mode, CC Switch incorrectly injected an `x-anthropic-billing-header` containing a dynamic `cch` value at the start of the `instructions` field. Because OpenAI prompt caching matches on a prefix, that header changing on every request stopped a large, otherwise stable system prompt from ever hitting the cache.

The cluster also holds unfixed variants: Azure OpenAI `max_tokens` not being converted to `max_completion_tokens`, producing a 400 (#3185); `supports_reasoning_effort()` using `strip_prefix("gpt-")` so a model with a custom prefix never triggers reasoning mapping (#5343); Claude Desktop routing not reusing the Session ID, lowering the cache rate (#5388); and `<thinking>` tags leaking into the body on the Codex chat-completions route (#6116).

Fix

Section titled “Fix”

  1. 区分两类代理：供应商配置里的 env proxy 只作用于 Claude Code / Codex 等 CLI 的请求流量，cc-switch 自身的内部请求（会话刷新、配额刷新）不会使用它，需要单独配置。

  2. 如果内部请求需要走代理：在 cc-switch 应用设置中开启并配置软件级「出站代理」，或配置系统代理。应用级出站代理优先，未设置时回退到系统代理。

  3. 针对 models.dev 定价同步报「Fetch is aborted」：该请求由 WebView 发出、只跟随系统代理，临时办法是设置系统代理（或 TUN）让 WKWebView 能经代理访问 models.dev；应用内代理对此无效，根治需等待官方把该 fetch 移到 Rust 后端的修复版本发布。

  4. 验证：`~/.cc-switch/model-pricing.json` 中的 `lastSyncError` 不再为 `Fetch is aborted`，且 `lastSyncAt` 出现时间戳，说明同步成功。

    ~/.cc-switch/model-pricing.json（同步成功后应看到 lastSyncAt 有值、lastSyncError 为空）
    {
      "lastSyncAt": "<时间戳>",
      "lastSyncError": null
    }

Affected Versions

Section titled “Affected Versions”
Toolcc-switch（应用自身的会话刷新与使用统计/定价同步功能）
Versionmodels.dev 同步失败确认存在于 3.20.0 且未修复；会话刷新走应用级出站代理/系统代理的行为自 3.13.0 起被用户观察到
PlatformsmacOSWindows

Source Issues

Section titled “Source Issues”

This page is distilled from 2 real issues

FAQ

Section titled “FAQ”
会话刷新走的是配置文件环境变量里的 proxy，还是软件里的出站 proxy？
根据维护者答复：会话刷新使用 cc-switch 的 HTTP 客户端，设置了应用级出站代理就走该代理，否则走系统代理；供应商 env 里的 proxy 设置不被使用。
如果配置文件设置了 env proxy，能否自动填充到软件单独设置的 proxy 处？
输入的 issue 中官方未回应该建议；目前两者是相互独立的配置，内部请求只认应用级出站代理或系统代理，需要手动单独设置。
我已在 cc-switch 里配置了代理，为什么 models.dev 定价同步还是失败？
定价同步在 WebView 内用浏览器原生 fetch 直连 models.dev，不经过 Rust 后端的全局代理客户端，WKWebView 只跟随系统代理，不读应用内代理配置。临时办法是设置系统代理；根治修复（把 fetch 挪到 Rust 后端）截至 v3.20.0 尚未发布。
「从 models.dev 导入」定价对话框也失败，是同一个问题吗？
是。导入对话框与自动同步共用同一条 fetch 路径，同样不走应用内代理，会报同样的 Fetch is aborted。
Section titled “Related problems”
<thinking> tags output as body text in Claude CodeChanging the provider's apiFormat from openai_chat to anthropic, if supported upstream, can bypass inline <thinking> leakage.1 issues mergedProvider API Key Overwritten: After Codex Restart or SwitchUpgrade to cc-switch v4.0.0 or later to prevent provider keys from overwriting each other and restore the official login state.6 issues mergedmissing field `call_id`DeepSeek Responses rejected a Codex delegated task entry missing `call_id`; the issue did not confirm a user-side fix.1 issues merged

This is an unofficial community wiki with no affiliation to the cc-switch authors or the project itself. Its content is compiled from the project's public GitHub issues. This site distributes no software.