feat(wren-ai-service): add support for new GPT-5 models in configuration files

Author: yichieh-lu

deployment/kustomizations/base/cm.yaml

@@ -75,7 +75,24 @@ data:
         n: 1
         seed: 0
         temperature: 0
-
+    - model: gpt-5-nano-2025-08-07
+      context_window_size: 1000000
+      kwargs:
+        max_tokens: 4096
+        n: 1
+        seed: 0
+    - model: gpt-5-mini-2025-08-07
+      context_window_size: 1000000
+      kwargs:
+        max_tokens: 4096
+        n: 1
+        seed: 0
+    - model: gpt-5-2025-08-07
+      context_window_size: 1000000
+      kwargs:
+        max_tokens: 4096
+        n: 1
+      seed: 0
     ---
     type: embedder
     provider: litellm_embedder

docker/config.example.yaml

@@ -24,7 +24,25 @@ models:
       n: 1
       seed: 0
       temperature: 0
-
+  - model: gpt-5-nano-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-mini-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+      
 ---
 type: embedder
 provider: litellm_embedder

wren-ai-service/poetry.lock

@@ -10,7 +10,7 @@ package-mode = false
 [tool.poetry.dependencies]
 python = ">=3.12.*, <3.13"
 fastapi = "^0.115.2"
-uvicorn = {extras = ["standard"], version = "^0.30.1"}
+uvicorn = {extras = ["standard"], version = "^0.29.0"} # Litellm requires uvicorn < 0.30.0
 python-dotenv = "^1.0.1"
 haystack-ai = "==2.7.0"
 openai = "^1.40.0"
@@ -33,8 +33,8 @@ pydantic-settings = "^2.5.2"
 google-auth = "^2.35.0"
 tiktoken = "^0.8.0"
 jsonschema = "^4.23.0"
-litellm = "^1.70.0"
-boto3 = "^1.35.90"
+litellm = {extras = ["proxy"], version = "^1.75.2"}
+boto3 = "^1.34.34" # Litellm requires boto3 = 1.34.34
 qdrant-client = "==1.11.0"
 
 [tool.poetry.group.dev.dependencies]

wren-ai-service/pyproject.toml

@@ -24,6 +24,24 @@ models:
       n: 1
       seed: 0
       temperature: 0
+  - model: gpt-5-nano-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-mini-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
 
 ---
 type: embedder

wren-ai-service/tools/config/config.example.yaml

@@ -24,6 +24,24 @@ models:
       n: 1
       seed: 0
       temperature: 0
+  - model: gpt-5-nano-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-mini-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
+  - model: gpt-5-2025-08-07
+    context_window_size: 1000000
+    kwargs:
+      max_tokens: 4096
+      n: 1
+      seed: 0
 
 ---
 type: embedder

wren-ai-service/tools/config/config.full.yaml

@@ -109,7 +109,7 @@ func askForGenerationModel() (string, error) {
 
 	prompt := promptui.Select{
 		Label: "Select an OpenAI's generation model",
-		Items: []string{"gpt-4.1", "gpt-4.1-mini", "gpt-4.1-nano"},
+		Items: []string{"gpt-4.1", "gpt-4.1-mini", "gpt-4.1-nano", "gpt-5", "gpt-5-mini", "gpt-5-nano"},
 	}
 
 	_, result, err := prompt.Run()
@@ -389,6 +389,9 @@ func getOpenaiGenerationModel() (string, bool) {
 			"gpt-4.1":      true,
 			"gpt-4.1-mini": true,
 			"gpt-4.1-nano": true,
+			"gpt-5":        true,
+			"gpt-5-mini":   true,
+			"gpt-5-nano":   true,
 		}
 		if !validModels[openaiGenerationModel] {
 			pterm.Error.Println("Invalid generation model", openaiGenerationModel)