Errors: Plugin DaemonInner Error or GPT 5 Nano Exist

[Jiaao2025]

Self Checks

• I have read the Contributing Guide and Language Policy.
• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report, otherwise it will be closed.
• 【中文用户 & Non English User】请使用英语提交，否则会被关闭 ：）
• Please do not modify this template :) and fill in all the required fields.

Dify version

1.9.1

Cloud or Self Hosted

Cloud

Steps to reproduce

I have experienced issues below all of the sudden, just by triggering my workflow.

req_id: 9e1e2daabe PluginInvokeError: {"args":null,"error_type":"PluginDaemonInnerError","message":"encountered an error: invalid character '\u003c' looking for beginning of value status: 503 Service Unavailable original response: \u003chtml\u003e"}

✔️ Expected Behavior

I should be expecting the workflow triggers as usual, while it suddenly keeps triggering this error

❌ Actual Behavior

It keeps flagging the internal server error and
PluginInvokeError: {"args":null,"error_type":"PluginDaemonInnerError","message":"encountered an error: invalid character '\u003c' looking for beginning of value status: 503 Service Unavailable original response: \u003chtml\u003e"}
or GPT does not exist

[dosubot]

Hi @Jiaao2025! I'm Dosu and I’m helping the dify team.

The 503 Service Unavailable and PluginDaemonInnerError with "invalid character '<'" happen when the plugin or model server returns an HTML error page instead of JSON, usually because the service is temporarily down or overloaded. This is a known issue in Dify Cloud v1.9.1 and is tracked in issue langgenius/dify#1542. The error handling logic in Dify surfaces these failures clearly in workflow logs and tracing, so checking those logs is the best way to pinpoint which node or plugin is failing.

For the "gpt-5-nano does not exist" error, this typically means the model is either unavailable, misconfigured, or not supported in your current environment. There was a recent fix (PR langgenius/dify#26421) that improves model list refresh when changing credentials, which helps prevent stale model lists from causing false "does not exist" errors. If you recently changed model provider credentials, try refreshing your model list or re-authenticating the provider.

Troubleshooting steps: verify the plugin/model provider status, refresh model lists after any credential changes, and check tracing logs for more details on which node is failing. If the issue persists, it may be due to a temporary outage or configuration problem with the plugin or model provider.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}