求助，codex无法连接CPA

问题描述：

codex一直报错
image1954×241 6.11 KB
image1572×194 2.46 KB
CPA能看到连续几个200，没有报错
image1755×500 80.4 KB
debug模式详细日志
image1909×283 44.9 KB

在cc-switch检测是正常的，在codex无法使用
在cherry studio使用OpenAI compatible接口访问GPT5.4完全正常，但用responses接口就空回
TUN模式，试了几个不同机场的节点，节点正常打开google，但codex还是不能用
真没找了，求佬友帮忙

网友解答：

---

--【壹】--：

我昨天也遇到了，时好时坏，不知道为啥

都是什么请求错误有请求id的

---

--【贰】--：

捞捞，有佬懂吗

---

--【叁】--：

之前我也出现过这个问题，忘了怎么弄的了

---

--【肆】--：

建议使用美国家宽代理 这样会稳定一些 不容易挂

---

--【伍】--：

问题解决，公益站那边挂掉之后，CPA不能很好的处理，仍然给出200，把公益站提供商关掉换官方认证文件就好了
另外建议佬友们用docker compose运行CPA

---

--【陆】--：

不行。。。。

---

--【柒】--：

不懂，等其他大佬，CCS看着没啥问题

---

--【捌】--：

这是cc-switch的

model_provider = "newapi" model = "gpt-5.4" model_reasoning_effort = "xhigh" requires_openai_auth = "true" startup_timeout_sec = "30" [model_providers] [model_providers.newapi] name = "NewAPI" base_url = "http://服务器ip:8317/v1" wire_api = "responses" requires_openai_auth = true [notice] [notice.model_migrations] "gpt-5.2" = "gpt-5.3-codex" "gpt-5.3-codex" = "gpt-5.4" [windows] sandbox = "elevated" [mcp_servers] [mcp_servers.grok-tavily-search] type = "stdio" command = "uvx" args = ["--from", "git+https://github.com/GuDaStudio/GrokSearch@grok-with-tavily", "grok-search"] [mcp_servers.grok-tavily-search.env] GROK_API_KEY = "sk-48" GROK_API_URL = "https://free.luciferai.cc/v1" GROK_MODEL = "grok-4.1-fast" TAVILY_API_KEY = "tvly-dev-H3" TAVILY_API_URL = "https://api.tavily.com"

这是CPA的：

# Server port port: 8317 # TLS settings for HTTPS. When enabled, the server listens with the provided certificate and key. tls: enable: false # Management API settings remote-management: # Whether to allow remote (non-localhost) management access. # When false, only localhost can access management endpoints (a key is still required). allow-remote: true # Management key. If a plaintext value is provided here, it will be hashed on startup. # All management requests (even from localhost) require this key. # Leave empty to disable the Management API entirely (404 for all /v0/management routes). secret-key: "$2a" # Disable the bundled management control panel asset download and HTTP route when true. disable-control-panel: false # GitHub repository for the management control panel. Accepts a repository URL or releases API URL. panel-github-repository: "https://github.com/router-for-me/Cli-Proxy-API-Management-Center" # Authentication directory (supports ~ for home directory) auth-dir: "/root/.cli-proxy-api" # API keys for authentication api-keys: - api # Enable debug logging debug: true # When true, disable high-overhead HTTP middleware features to reduce per-request memory usage under high concurrency. commercial-mode: false # When true, write application logs to rotating files instead of stdout logging-to-file: true # Maximum total size (MB) of log files under the logs directory. When exceeded, the oldest log # files are deleted until within the limit. Set to 0 to disable. logs-max-total-size-mb: 10 # When false, disable in-memory usage statistics aggregation usage-statistics-enabled: true # When true, unprefixed model requests only use credentials without a prefix (except when prefix == model name). force-model-prefix: false # Number of times to retry a request. Retries will occur if the HTTP response code is 403, 408, 500, 502, 503, or 504. request-retry: 3 # Maximum wait time in seconds for a cooled-down credential before triggering a retry. max-retry-interval: 40 # Quota exceeded behavior quota-exceeded: switch-project: true # Whether to automatically switch to another project when a quota is exceeded switch-preview-model: true # Whether to automatically switch to a preview model when a quota is exceeded # Routing strategy for selecting credentials when multiple match. routing: strategy: "fill-first" # round-robin (default), fill-first # When true, enable authentication for the WebSocket API (/v1/ws). ws-auth: false # Streaming behavior (SSE keep-alives + safe bootstrap retries). # streaming: # keepalive-seconds: 15 # Default: 0 (disabled). <= 0 disables keep-alives. # bootstrap-retries: 1 # Default: 0 (disabled). Retries before first byte is sent. openai-compatibility: - name: MS base-url: https://api-inference.modelscope.cn/v1 api-key-entries: - api-key: 8 models: - name: Qwen/Qwen3.5-122B-A10B alias: "Qwen3.5-122B-A10B" - name: Qwen/Qwen3.5-27B alias: "Qwen3.5-27B" - name: Qwen/Qwen3.5-397B-A17B alias: "Qwen3.5-397B-A17B" - name: infini base-url: https://cloud.infini-ai.com/maas/coding/v1 api-key-entries: - api-key: sk-cp- - api-key: sk-cp- models: - name: kimi-k2.5 alias: "" - name: glm-5 alias: "" claude-api-key: - api-key: sk-h base-url: https://yyds.215.im proxy-url: "" models: - name: glm-5 alias: "" - name: claude-opus-4-6 alias: "" codex-api-key: - api-key: sk-9 base-url: https://ai.qaq.al websockets: true proxy-url: "" models: - name: gpt-5.2 alias: "" - name: gpt-5.4 alias: "" - name: gpt-5.3-codex alias: "" prefix: ggb priority: 100 excluded-models: - '*' - api-key: sk-G base-url: https://imsnake.dart.us.ci websockets: true proxy-url: "" models: - name: qwen3.5-plus-thinking alias: "" - name: qwen3.5-397b-a17b-thinking alias: "" - name: gpt-5.3-codex alias: "" - name: gpt-5.4 alias: "" priority: 100 - api-key: sk-9 priority: 99 base-url: https://api.268526.eu.cc proxy-url: "" models: - name: gpt-5.3-codex alias: "" - name: gpt-5.4 alias: "" websockets: true excluded-models: - '*' - api-key: sk-K priority: 99 base-url: https://sakuradori.dpdns.org proxy-url: "" models: - name: glm-5 alias: "" - name: gpt-5.3-codex alias: "" - name: gpt-5.4 alias: "" - name: Qwen3.5-397B-A17B alias: "" websockets: true - api-key: sk-b priority: 100 base-url: https://api.nih.cc websockets: true proxy-url: "" models: - name: z-ai/glm5 alias: glm-5 - name: anthropic/claude-sonnet-4.6-thinking alias: claude-sonnet-4.6 - name: gpt-5.4 alias: "" - name: gpt-5.3-codex alias: "" - name: moonshotai/kimi-k2.5 alias: kimi-k2.5 - api-key: sk-G priority: 99 prefix: wind base-url: https://windhub.cc websockets: true proxy-url: "" models: - name: gpt-5.4 alias: "" - name: gpt-5.3-codex alias: "" excluded-models: - '*' - api-key: sk-b base-url: https://china.184772.xyz websockets: true proxy-url: "" models: - name: gpt-5.3-codex alias: "" - name: gpt-5.4 alias: "" - name: claude-opus-4-6 alias: "" priority: 99 - api-key: sk-i base-url: https://lx.lxsummer.cloud proxy-url: "" models: - name: gemini-3.1-pro alias: "" - name: 「Summer」gpt-5.4 alias: gpt-5.4 priority: 99 - api-key: sk-0 priority: 99 prefix: ice base-url: https://ice.v.ua websockets: true proxy-url: "" models: - name: claude-opus-4-6 alias: "" - name: gpt-5.3-codex alias: "" - name: gpt-5.4 alias: "" - name: gpt-5.4-pro alias: "" ampcode: upstream-url: "" upstream-api-key: sgamp_user_0 restrict-management-to-localhost: false model-mappings: [] force-model-mappings: true oauth-model-alias: antigravity: - name: gemini-3-flash alias: claude-haiku-4-5 fork: true - name: gemini-3-pro-image alias: gemini-3-pro-image-preview - name: gemini-3-pro-high alias: gemini-3-pro-preview - name: gemini-3.1-pro-high alias: gemini-3.1-pro-preview - name: gemini-3-flash alias: gemini-3-flash-preview fork: true pprof: enable: false addr: 127.0.0.1:8316 error-logs-max-files: 10 gemini-api-key: - api-key: sk-C prefix: bh base-url: https://x666.me models: - name: gemini-3-flash-preview alias: "" priority: 100 - api-key: AIza models: - name: gemini-3-flash-preview alias: "" - api-key: AIza models: - name: gemini-3-flash-preview alias: "" - api-key: AIza models: - name: gemini-3-flash-preview alias: "" - api-key: sk-4 base-url: https://ai.hybgzs.com/gemini models: - name: gemini-3.1-pro-preview alias: "" - name: gemini-3-flash-preview alias: "" priority: 100 - api-key: sk-D base-url: https://ai.xiangcao.de models: - name: gemini-3.1-pro-preview alias: "" - name: gemini-3-flash-preview alias: "" - name: gemini-3.1-flash-lite-preview alias: "" priority: 99 - api-key: sk-T base-url: https://elysiver.h-e.top models: - name: gemini-3.1-pro-preview alias: "" - name: gemini-3-flash-preview alias: "" priority: 99 - api-key: sk-k priority: 98 base-url: https://hello.xhyun.eu.cc models: - name: gemini-2.5-pro alias: "" - name: gemini-3-flash-preview alias: "" excluded-models: - '*' - api-key: sk-X base-url: https://wzw.pp.ua models: - name: gemini-3.1-pro-preview alias: "" - name: gemini-2.5-pro alias: "" - name: gemini-3-flash-preview alias: "" - name: gemini-3.1-flash-lite-preview alias: "" priority: 99 - api-key: sk-t priority: 99 base-url: https://new.api-test.us.ci models: - name: gemini-3-flash-preview alias: "" - name: gemini-3.1-flash-image-preview alias: "" - name: gemini-3.1-pro-preview alias: "" - api-key: sk-V priority: 100 base-url: https://aiapi.exe.xyz models: - name: gpt-5.4 alias: "" prefix: test excluded-models: - '*' - api-key: sk-E base-url: https://ai.prism.uno models: - name: gemini-3.1-pro-preview alias: "" - name: gemini-3-flash-preview alias: "" oauth-excluded-models: vertex: - gemini-2.5-pro - gemini-2.5-flash - gemini-2.5-flash-lite - gemini-3-pro-preview - gemini-3-flash-preview - gemini-3-pro-image-preview - imagen-4.0-generate-001 - imagen-3.0-generate-002 - imagen-4.0-ultra-generate-001 - imagen-3.0-fast-generate-001 - imagen-4.0-fast-generate-001 streaming: keepalive-seconds: 30 bootstrap-retries: 3 nonstream-keepalive-interval: 10

---

--【玖】--：

config文件看看

---

--【拾】--：

佬友，你有没有加域名？还是用其他方法解决的，我发现就responses api不能用，而且是近几天出现的

---

--【拾壹】--：

试试
PixPin_2026-04-01_18-38-55920×300 20.3 KB

---

--【拾贰】--：

更新了吧，要改下配置

---

--【拾叁】--：

看看使用统计的token是不是都是0，cpa如果接有些公益站而不是认证文件是会出现这样的问题

---

--【拾肆】--：

什么配置啊

---

--【拾伍】--：

根本没有请求事件统计，这也是我不理解的