FastGPT Model serving and inference Issue List

This page collects the 41 published documents about self-hosted inference services and model integration, grouped by symptom, so a document can be found directly from what the error looks like.

Three symptoms that belong to this stage

The inference endpoint is reachable but calls fail
Retrieval behaves differently after changing an embedding or rerank model
Channel and credential setup in the model proxy layer

If none of the three match, go back to the deployment and environment issue landscape and narrow down again.

The general order for this stage

Take the full error from the backend service log for this stage's component, including the component name and error code
Verify from inside the deployment that the component can be reached on its own, ruling out network and permission causes
Check the component version against the main service version
Follow the entry below whose symptom is closest, then repeat the same operation to verify

Published documents (41)

Document	Area
Add Custom SiliconCloud Models to FastGPT	deploy
Add MiniMax Models to FastGPT Self-Hosted Deployments	deploy
Add Multilingual Model Provider Display Names	model
Add Valid Models to FastGPT Provider Presets	model
Add a Model to an Existing FastGPT Provider	model
Add a New Model Provider to FastGPT	model
Add and Maintain FastGPT Model Presets	model
Add and Use Ollama Models in FastGPT	deploy
Clarify FastGPT Dataset Processing Model Roles	tutorial
Deploy ChatGLM2 Models via FastGPT Source Code	deploy
Deploy M3E Large Model API for FastGPT	deploy
Deploy Qwen Models Using Xinference for FastGPT	deploy
Deploy and Use Custom ChatGLM2 and M3E Models on FastGPT	deploy
Download BGE Rerank Model Code for FastGPT	deploy
Download BGE Reranker Models for FastGPT Deployment	deploy
Enable and Map FastGPT AI Models	deploy
Fix Non-Native Model Question Classification Issues	deploy
Install Ollama for FastGPT Self-Hosted Deployments	deploy
Install Ollama on Host Machine for FastGPT	deploy
Install and Deploy Xinference for FastGPT Custom Models	deploy
Integrate Local Models For FastGPT Via One API	deploy
Integrate Local Models via Xinference with FastGPT	deploy
Integrate Qwen Local Model with FastGPT	deploy
Launch Xinference Models via Command Line Interface	deploy
Manage and Access FastGPT Model Call Logs	deploy
Match FastGPT Channel IDs to AIProxy Protocols	model
Replace OneAPI with AI Proxy for FastGPT	deploy
Resolve FastGPT Empty Model Response Errors	deploy
Set Up Direct Ollama Integration With FastGPT	deploy
Set Up FastGPT AI Model Basic Settings	tutorial
Set Up FastGPT AIProxy Protocol Entries	model
Set Up FastGPT Custom Model Request URLs	deploy
Set Up M3E Embedding Models in FastGPT	deploy
Set Up Ollama AI Proxy Integration for FastGPT	deploy
Set Up and Manage FastGPT Model Channels	deploy
Test Custom ChatGLM2 Models on FastGPT	deploy
Test Custom ChatGLM2 and M3E Models on FastGPT	deploy
Troubleshoot FastGPT Model Availability Errors	deploy
Understand FastGPT Model Preset Directory Structures	model
Validate Ollama Connectivity for FastGPT Deployments	deploy
Validate Updated FastGPT Model Preset Changes	model

What this list does not cover

Entries cover cases that can be reproduced publicly, grouped by symptom. The following need separate confirmation:

A symptom caused by several factors at once, which needs the order above to rule them out one by one
The same symptom caused by configuration specific to the commercial edition
Cases coupled to a particular infrastructure environment that cannot be reproduced in a standard deployment

Keep reading

References

If the problem is still not located

The entries above cover cases that can be reproduced from public information. If the problem depends on configuration details of a specific deployment, or needs runtime logs to confirm, contact sales for deployment-stage support; the cloud service can be used directly without handling environment dependencies.

Contact sales: support for self-hosting and upgrades
Get started: use the cloud service and skip environment setup
Pricing: compare what the cloud and self-hosted forms cover