Repository navigation
GPT 5.5 Custom Provider Configuration fails to automatically locate /v1/responses endpoint #29325
Description
Description
"provider": {
"gpt_5_5": {
"name": "gpt_5_5",
"npm": "@ai-sdk/openai-compatible",
"options": {
"baseURL": "$BASE_URL/v1"
},
"models": {
"gpt-5.5": {
"name": "gpt-5.5"
}
}
}
}
This configuration is showing an error：
Invalid param: Function tools with reasoning_effort are not supported for gpt-5.5 in /v1/chat/completions. Please use /v1/responses instead.
I cannot find how to configure the modified endpoint. Please add a configurable endpoint.
Plugins
No response
OpenCode version
1.15.10
Steps to reproduce
- Configure a custom GPT 5.5 provider
- Select GPT 5.5 and say hello
- Reproduce the error
Screenshot and/or share link
No response
Operating System
MacBook Pro, Apple M4, macOS Tahoe 26.5
Terminal
No response
Metadata
Metadata
Assignees
Labels
No labels
Activity
github-actions commentedon May 26, 2026
This issue might be a duplicate of existing issues. Please check:
@ai-sdk/openai-compatiblewithgpt-5.5— reports a different symptom (mid-stream EOF) but the same provider setup and version. The root cause may overlap.blankprojects commentedon May 26, 2026
After much trial and error with the large model, I discovered that changing "npm": "@ai-sdk/openai-compatible" to "npm": "@ai-sdk/openai" is effective.