PluginInvokeError: {"args":{},"error_type":"NotFoundError","message":"Error code: 404 - {'error': {'message': 'page not found'}}"}

[Moiyu1998]

Self Checks

• [x] This is only for bug report, if you would like to ask a question, please head to Discussions.
• [x] I have searched for existing issues search for existing issues, including closed ones.
• [x] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [x] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• [x] Please do not modify this template :) and fill in all the required fields.

Dify version

1.1.3

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

When calling the model, no result was returned. Checking the logs revealed an error: core.plugin.manager.exc.PluginInvokeError: PluginInvokeError: {"args":{},"error_type":"NotFoundError","message":"Error code: 404 - {'error': {'message': 'page not found'}}"}

This situation has occurred with many models, such as GPT-4.1 and Claude 3.7, among others.

✔️ Expected Behavior

No response

❌ Actual Behavior

No response

[dosubot[bot]]

The NotFoundError you're encountering during plugin invocation in Dify can be caused by several factors. Here are some potential solutions:

1. Connectivity Issues: Ensure that port 443 is not blocked by your firewall, as this can prevent access to necessary external sources like GitHub and Python package repositories ^[1].

2. Timeout Settings: In Dify v1.1.2 (self-hosted), a 'plugin not found' error can occur due to timeout settings. You might need to modify the timeout settings in the docker-compose.middleware.yaml file to resolve this issue ^[2].

3. Model Compatibility: As of Dify v1.2.0, there is no official support for Azure OpenAI's GPT-4.1 series models. Attempts to manually add these models with OpenAI-API-compatible settings may result in errors. Official support for these models will be added in a future update ^[3].

4. Plugin Installation Issues: If the error persists, it might be due to failed plugin installations. You can try clearing plugin data from the database, clearing plugin directories, restarting the plugin daemon service, and reinstalling the plugins ^[4].

By addressing these areas, you should be able to resolve the NotFoundError and successfully invoke the models.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}

[Moiyu1998]

More detailed error information as above.

[Moiyu1998]

It seems that the error occurred when requesting the /plugin/dc611882-13de-49f0-8fa6-f2a23a047266/dispatch/llm/num_tokens endpoint.

[Moiyu1998]

After updating the OpenAI plugin, the GPT-4.1 model returned to normal, but the Claude model is still reporting this error.

[Moiyu1998]

[Moiyu1998]

@crazywoola Please try using the Claude 3.7 model to see if the issue can be reproduced.

[yanyuw]

I have same problem. And it still reporting this error with Claude 3.7 model

[Moiyu1998]

It seems that the issue is likely related to the Claude plugin.

[crazywoola]

PLUGIN_DIFY_INNER_API_KEY=QaHbTe77CtuXmsfyhR7+vRjI/+XbV1AaFy691iy+kGDv2Jvy0/eAh8Y1
PLUGIN_DIFY_INNER_API_URL=http://api:500

See this thread.

[Moiyu1998]

PLUGIN_DIFY_INNER_API_KEY=QaHbTe77CtuXmsfyhR7+vRjI/+XbV1AaFy691iy+kGDv2Jvy0/eAh8Y1
PLUGIN_DIFY_INNER_API_URL=http://api:500

See this thread.

The environment variables have been changed, but this error is still being reported.

[Moiyu1998]

@crazywoola @laipz8200 I printed the request body in the logs, which is the same as the one shown in the image above. Then, I used the curl command inside the container to send this request directly, and it returned the same error. I have done my best to help you locate the issue. I have noticed that other users have also reported this issue, but the issues were closed without resolving the problem. This appears to be a bug. Please investigate and resolve this issue, as the Claude model series is currently unusable.

[Moiyu1998]

After I updated Dify to version 1.3.0, this issue was resolved.