feat: support lingyiwanwu client

Argcfile.sh

@@ -91,7 +91,8 @@ OPENAI_COMPATIBLE_PLATFORMS=( \
   octoai,meta-llama-3-8b-instruct,https://text.octoai.run/v1 \
   perplexity,llama-3-8b-instruct,https://api.perplexity.ai \
   together,meta-llama/Llama-3-8b-chat-hf,https://api.together.xyz/v1 \
-  zhipuai,glm-4,https://open.bigmodel.cn/api/paas/v4 \
+  zhipuai,glm-4-0520,https://open.bigmodel.cn/api/paas/v4 \
+  lingyiwanwu,yi-large,https://api.lingyiwanwu.com/v1 \
 )
 
 # @cmd Chat with any LLM api 

config.example.yaml

@@ -112,7 +112,7 @@ clients:
     api_key: xxx                                      # ENV: {client}_API_KEY
     patches:
       '.*':                                           
-        chat_completions_body:                        # Override safetySettings for all models
+        chat_completions_body:
           safetySettings:
             - category: HARM_CATEGORY_HARASSMENT
               threshold: BLOCK_NONE
@@ -178,7 +178,7 @@ clients:
     adc_file: <path-to/gcloud/application_default_credentials.json> 
     patches:
       'gemini-.*':
-        chat_completions_body:                        # Override safetySettings for all gemini models
+        chat_completions_body:
           safetySettings:
             - category: HARM_CATEGORY_HARASSMENT
               threshold: BLOCK_ONLY_HIGH

models.yaml

@@ -588,6 +588,41 @@
       output_price: 7
       supports_vision: true
 
+- platform: lingyiwanwu
+  # docs:
+  #   - https://platform.lingyiwanwu.com/docs#%E6%A8%A1%E5%9E%8B
+  #   - https://platform.lingyiwanwu.com/docs#%E8%AE%A1%E8%B4%B9%E5%8D%95%E5%85%83
+  models:
+  - name: yi-large 
+    max_input_tokens: 32768
+    input_price: 2.8
+    output_price: 2.8
+  - name: yi-medium 
+    max_input_tokens: 16384
+    input_price: 0.35
+    output_price: 0.35
+  - name: yi-vision
+    max_input_tokens: 4096
+    input_price: 0.84
+    output_price: 0.84
+    supports_vision: true
+  - name: yi-medium-200k
+    max_input_tokens: 200000
+    input_price: 1.68
+    output_price: 1.68
+  - name: yi-spark
+    max_input_tokens: 16384
+    input_price: 0.14
+    output_price: 0.14
+  - name: yi-large-rag
+    max_input_tokens: 16384
+    input_price: 3.5
+    output_price: 3.5
+  - name: yi-large-turbo
+    max_input_tokens: 16384
+    input_price: 1.68
+    output_price: 1.68
+
 - platform: anyscale
   # docs:
   #   - https://docs.anyscale.com/endpoints/text-generation/supported-models/meta-llama-Meta-Llama-3-8B-Instruct

src/client/mod.rs

@@ -45,7 +45,7 @@ register_client!(
     (qianwen, "qianwen", QianwenConfig, QianwenClient),
 );
 
-pub const OPENAI_COMPATIBLE_PLATFORMS: [(&str, &str); 12] = [
+pub const OPENAI_COMPATIBLE_PLATFORMS: [(&str, &str); 13] = [
     ("anyscale", "https://api.endpoints.anyscale.com/v1"),
     ("deepinfra", "https://api.deepinfra.com/v1/openai"),
     ("deepseek", "https://api.deepseek.com"),
@@ -58,4 +58,5 @@ pub const OPENAI_COMPATIBLE_PLATFORMS: [(&str, &str); 12] = [
     ("perplexity", "https://api.perplexity.ai"),
     ("together", "https://api.together.xyz/v1"),
     ("zhipuai", "https://open.bigmodel.cn/api/paas/v4"),
+    ("lingyiwanwu", "https://api.lingyiwanwu.com/v1"),
 ];