jiazhizhong/higress
1
0
Fork 0
mirror of https://github.com/alibaba/higress.git synced 2026-03-06 17:40:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat(provider): add support for Grok provider in AI proxy (#2713)

Browse Source 
Co-authored-by: 韩贤涛 <601803023@qq.com>
This commit is contained in:
aias00
2025-08-07 17:22:47 +08:00
committed by GitHub
parent b69e3a8f30
commit a1bf1ff009
6 changed files with 354 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files

60
test/e2e/conformance/tests/go-wasm-ai-proxy.go
View File

@@ -1023,6 +1023,66 @@ data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"role":"assistant
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"role":"assistant","content":"USER: \n你好\n"},"finish_reason":"stop","logprobs":null}],"model":"dify","object":"chat.completion.chunk","usage":{"prompt_tokens":9,"completion_tokens":1,"total_tokens":10}}
`),
},
},
},
{
Meta: http.AssertionMeta{
TestCaseName: "grok case 1: non-streaming request",
CompareTarget: http.CompareTargetResponse,
},
Request: http.AssertionRequest{
ActualRequest: http.Request{
Host: "api.x.ai",
Path: "/v1/chat/completions",
Method: "POST",
ContentType: http.ContentTypeApplicationJson,
Body: []byte(`{"model":"gpt-3","messages":[{"role":"user","content":"你好，你是谁？"}],"stream":false}`),
},
},
Response: http.AssertionResponse{
ExpectedResponse: http.Response{
StatusCode: 200,
ContentType: http.ContentTypeApplicationJson,
Body: []byte(`{"id":"chatcmpl-llm-mock","choices":[{"index":0,"message":{"role":"assistant","content":"你好，你是谁？"},"finish_reason":"stop","logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion","usage":{"prompt_tokens":9,"completion_tokens":1,"total_tokens":10}}`),
},
},
},
{
Meta: http.AssertionMeta{
TestCaseName: "grok case 2: streaming request",
CompareTarget: http.CompareTargetResponse,
},
Request: http.AssertionRequest{
ActualRequest: http.Request{
Host: "api.x.ai",
Path: "/v1/chat/completions",
Method: "POST",
ContentType: http.ContentTypeApplicationJson,
Body: []byte(`{"model":"gpt-3","messages":[{"role":"user","content":"你好，你是谁？"}],"stream":true}`),
},
},
Response: http.AssertionResponse{
ExpectedResponse: http.Response{
StatusCode: 200,
ContentType: http.ContentTypeTextEventStream,
Body: []byte(`data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"你"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"好"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"，"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"你"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"是"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"谁"},"finish_reason":null,"logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: {"id":"chatcmpl-llm-mock","choices":[{"index":0,"delta":{"content":"？"},"finish_reason":"stop","logprobs":null}],"created":10,"model":"grok-beta","object":"chat.completion.chunk","usage":null}
data: [DONE]
`),
},
},

108
test/e2e/conformance/tests/go-wasm-ai-proxy.yaml
View File

@@ -334,6 +334,25 @@ spec:
port:
number: 3000
---
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
name: wasmplugin-ai-proxy-grok
namespace: higress-conformance-ai-backend
spec:
ingressClassName: higress
rules:
- host: "api.x.ai"
http:
paths:
- pathType: Prefix
path: "/"
backend:
service:
name: llm-mock-service
port:
number: 3000
---
apiVersion: extensions.higress.io/v1alpha1
kind: WasmPlugin
metadata:
@@ -349,8 +368,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': 360gpt-turbo
'*': 360gpt-pro
"gpt-3": 360gpt-turbo
"*": 360gpt-pro
type: ai360
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-ai360
@@ -359,8 +378,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': baichuan2-13b-chat-v1
'*': baichuan-7b-v1
"gpt-3": baichuan2-13b-chat-v1
"*": baichuan-7b-v1
type: baichuan
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-baichuan
@@ -369,8 +388,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': ernie-3.5-8k
'*': ernie-3.5-8k
"gpt-3": ernie-3.5-8k
"*": ernie-3.5-8k
type: baidu
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-baidu
@@ -379,8 +398,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': deepseek-reasoner
'*': deepseek-chat
"gpt-3": deepseek-reasoner
"*": deepseek-chat
type: deepseek
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-deepseek
@@ -389,7 +408,7 @@ spec:
apiTokens:
- fake_token
modelMapping:
'*': fake_doubao_endpoint
"*": fake_doubao_endpoint
type: doubao
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-doubao
@@ -398,8 +417,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': cohere-command-r-08-2024
'*': Phi-3.5-MoE-instruct
"gpt-3": cohere-command-r-08-2024
"*": Phi-3.5-MoE-instruct
type: github
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-github
@@ -408,8 +427,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': llama3-8b-8192
'*': llama-3.1-8b-instant
"gpt-3": llama3-8b-8192
"*": llama-3.1-8b-instant
type: groq
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-groq
@@ -418,9 +437,9 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': abab6.5s-chat
'gpt-4': abab6.5g-chat
'*': abab6.5t-chat
"gpt-3": abab6.5s-chat
"gpt-4": abab6.5g-chat
"*": abab6.5t-chat
type: minimax
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-minimax-v2-api
@@ -429,9 +448,9 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': abab6.5s-chat
'gpt-4': abab6.5g-chat
'*': abab6.5t-chat
"gpt-3": abab6.5s-chat
"gpt-4": abab6.5g-chat
"*": abab6.5t-chat
type: minimax
minimaxApiType: pro
minimaxGroupId: 1
@@ -442,8 +461,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': mistral-tiny
'*': mistral-large-latest
"gpt-3": mistral-tiny
"*": mistral-large-latest
type: mistral
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-mistral
@@ -452,10 +471,10 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': qwen-turbo
'gpt-35-turbo': qwen-plus
'gpt-4-*': qwen-max
'*': qwen-turbo
"gpt-3": qwen-turbo
"gpt-35-turbo": qwen-plus
"gpt-4-*": qwen-max
"*": qwen-turbo
type: qwen
qwenEnableCompatible: true
ingress:
@@ -465,10 +484,10 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': qwen-turbo
'gpt-35-turbo': qwen-plus
'gpt-4-*': qwen-max
'*': qwen-turbo
"gpt-3": qwen-turbo
"gpt-35-turbo": qwen-plus
"gpt-4-*": qwen-max
"*": qwen-turbo
type: qwen
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-qwen
@@ -477,8 +496,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': step-1-8k
'*': step-1-32k
"gpt-3": step-1-8k
"*": step-1-32k
type: stepfun
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-stepfun
@@ -487,8 +506,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': meta-llama/Meta-Llama-3-8B-Instruct-Turbo
'*': meta-llama/Llama-3-8b-chat-hf
"gpt-3": meta-llama/Meta-Llama-3-8B-Instruct-Turbo
"*": meta-llama/Llama-3-8b-chat-hf
type: together-ai
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-together-ai
@@ -497,8 +516,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': Yi-Medium
'*': Yi-Large
"gpt-3": Yi-Medium
"*": Yi-Large
type: yi
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-yi
@@ -507,8 +526,8 @@ spec:
apiTokens:
- fake_token
modelMapping:
'gpt-3': glm-4-plus
'*': glm-4-long
"gpt-3": glm-4-plus
"*": glm-4-long
type: zhipuai
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-zhipuai
@@ -517,9 +536,20 @@ spec:
apiTokens:
- fake_token
modelMapping:
'*': dify
"*": dify
type: dify
botType: Completion
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-dify
url: file:///opt/plugins/wasm-go/extensions/ai-proxy/plugin.wasm
- config:
provider:
apiTokens:
- fake_token
modelMapping:
"gpt-3": grok-beta
"gpt-4": grok-beta
"*": grok-beta
type: grok
ingress:
- higress-conformance-ai-backend/wasmplugin-ai-proxy-grok
url: file:///opt/plugins/wasm-go/extensions/ai-proxy/plugin.wasm