{
"modifiedIndex": 50332,
"value": {
"plugins": {
"ai-proxy-multi": {
"instances": [
{
"provider": "openai-compatible",
"auth": {
"header": {
"Authorization": "Bearer sk-21784d812fbc5d2ca3d3cbe02c9201b06cbad1372c237ecc"
}
},
"options": {
"model": "/models/Qwen3-Reranker-0.6B"
},
"weight": 100,
"name": "llm-instance-0",
"override": {
"endpoint": "http://172.32.150.72:8008"
}
}
],
"fallback_strategy": [
"rate_limiting",
"http_429",
"http_5xx"
]
},
"limit-conn": {
"key_type": "var",
"rejected_code": 503,
"rejected_msg": "Requests are too frequent, please try again later.",
"default_conn_delay": 0.1,
"conn": 11,
"burst": 0,
"key": "server_addr"
},
"proxy-rewrite": {
"regexUri": [
"^/.*",
"$0"
],
"empty": false
}
},
"create_time": 1775733072,
"timeout": {
"connect": 3,
"read": 180,
"send": 10
},
"id": "4850-1",
"plugin_config_id": "2",
"update_time": 1776081582,
"enable_websocket": false,
"methods": [
"POST"
],
"uri": "/*",
"priority": 3
},
"key": "/apisix/routes/4850-1",
"createdIndex": 50314
}
curl http://172.32.153.143:49080/v1/rerank -X POST -H "Content-Type:application/json;charset=UTF-8" \
-H "Authorization: d2zy6LfRgI" \
-d '{
"model": "rerank1",
"query": "What is the capital of France?",
"documents": [
"The capital of Brazil is Brasilia.",
"The capital of France is Paris.",
"Horses and cows are both animals"
],
"top_n":2
}'
Description
I have rerank server,the request url is
http://172.32.150.72:8008/v1/rerankandhttp://172.32.150.72:8008/score,I want to use ai-multi-proxy plugin ,because the endpoint must config like http://ip:port/pathhow to proxy both
http://172.32.150.72:8008/v1/rerankandhttp://172.32.150.72:8008/scoremy route config is
when request(the model
refank1has been transformed to /models/Qwen3-Reranker-0.6B)response
{"detail":"Not Found"}