Merge pull request #245 from MacPaw/feat/o1

Feat/o1

Author: nezhyborets

Demo/DemoChat/Sources/AssistantStore.swift

@@ -93,7 +93,7 @@ public final class AssistantStore: ObservableObject {
                         return nil
                     }
                 }
-                let fileIds = result.toolResources.codeInterpreter?.fileIds ?? []
+                let fileIds = result.toolResources?.codeInterpreter?.fileIds ?? []
 
                 assistants.append(Assistant(id: result.id, name: result.name ?? "", description: result.description, instructions: result.instructions, codeInterpreter: codeInterpreter, fileSearch: fileSearch, fileIds: fileIds, functions: functions))
             }

Sources/OpenAI/Public/Models/Models/Models.swift

@@ -10,6 +10,16 @@ public typealias Model = String
 
 public extension Model {
     // Chat Completion
+    
+    // o1 series
+    // The o1 series of models are trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.
+    
+    /// `o1`: reasoning model designed to solve hard problems across domains
+    static let o1 = "o1"
+    
+    /// `o1-mini`: fast and affordable reasoning model for specialized tasks
+    static let o1_mini = "o1-mini"
+    
     // GPT-4
 
     /// `gpt-4o`, currently the most advanced, multimodal flagship model that's cheaper and faster than GPT-4 Turbo.