Merge branch 'main' into nvirdy/verbosity

Author: neelvirdy

README.md

@@ -1159,11 +1159,14 @@ Models are represented as a typealias `typealias Model = String`.
 
 ```swift
 public extension Model {
+    static let gpt5_1 = "gpt-5.1"
+    static let gpt5_1_chat_latest = "gpt-5.1-chat-latest"
+
     static let gpt5 = "gpt-5"
     static let gpt5_mini = "gpt-5-mini"
     static let gpt5_nano = "gpt-5-nano"
     static let gpt5_chat = "gpt-5-chat"
-    
+
     static let gpt4_1 = "gpt-4.1"
     static let gpt4_1_mini = "gpt-4.1-mini"
     static let gpt4_1_nano = "gpt-4.1-nano"

Sources/OpenAI/Public/Models/ChatQuery.swift

@@ -15,7 +15,7 @@ public struct ChatQuery: Equatable, Codable, Streamable, Sendable {
     public let messages: [Self.ChatCompletionMessageParam]
     /// Model ID used to generate the response, like `gpt-4o` or `o3`. OpenAI offers a wide range of models with different capabilities, performance characteristics, and price points. Refer to the [model guide](https://platform.openai.com/docs/models) to browse and compare available models.
     public let model: Model
-    /// Constrains effort on reasoning for [reasoning models](https://platform.openai.com/docs/guides/reasoning). Currently supported values are minimal, low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+    /// Constrains effort on reasoning for [reasoning models](https://platform.openai.com/docs/guides/reasoning). Currently supported values are none, minimal, low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
     ///
     /// - Note: o-series models only
     public let reasoningEffort: ReasoningEffort?
@@ -922,6 +922,7 @@ public struct ChatQuery: Equatable, Codable, Streamable, Sendable {
     }
 
     public enum ReasoningEffort: Codable, Equatable, Sendable {
+        case none
         case minimal
         case low
         case medium
@@ -935,6 +936,8 @@ public struct ChatQuery: Equatable, Codable, Streamable, Sendable {
         public func encode(to encoder: any Encoder) throws {
             var container = encoder.singleValueContainer()
             switch self {
+            case .none:
+                try container.encode("none")
             case .minimal:
                 try container.encode("minimal")
             case .low:
@@ -952,6 +955,8 @@ public struct ChatQuery: Equatable, Codable, Streamable, Sendable {
             let container = try decoder.singleValueContainer()
             let rawValue = try container.decode(String.self)
             switch rawValue {
+            case "none":
+                self = .none
             case "minimal":
                 self = .minimal
             case "low":

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

@@ -53,6 +53,14 @@ public extension Model {
     @available(*, deprecated, message: "On April 14th, 2025, developers were notified that the gpt-4.5-preview model is deprecated and will be removed from the API in the coming months. Recommended replacement: gpt-4.1")
     static let gpt4_5_preview = "gpt-4.5-preview"
 
+    // GPT-5.1
+
+    /// `gpt-5.1` Enhanced version of GPT-5 with improved reasoning and performance
+    static let gpt5_1 = "gpt-5.1"
+
+    /// `gpt-5.1-chat-latest` Latest GPT-5.1 model optimized for chat interactions
+    static let gpt5_1_chat_latest = "gpt-5.1-chat-latest"
+
     // GPT-5
 
     /// `gpt-5` OpenAI's best AI system with significant leap in intelligence, designed for logic and multi-step tasks with deep reasoning
@@ -261,7 +269,7 @@ public extension Model {
             // reasoning
             .o4_mini, o3, o3_mini, .o1,
             // flagship
-            .gpt5, .gpt5_mini, .gpt5_nano, .gpt5_chat, .gpt4_1, .gpt4_o, .gpt_4o_audio_preview, chatgpt_4o_latest,
+            .gpt5, .gpt5_mini, .gpt5_nano, .gpt5_chat, .gpt5_1, .gpt5_1_chat_latest, .gpt4_1, .gpt4_o, .gpt_4o_audio_preview, chatgpt_4o_latest,
             // cost-optimized
             .gpt4_1_mini, .gpt4_1_nano, .gpt4_o_mini, .gpt_4o_mini_audio_preview,
             // tool-specific
@@ -274,7 +282,7 @@ public extension Model {
             // reasoning
             .o4_mini, .o3, .o3_mini, .o1, .o1_pro,
             // flagship
-            .gpt5, .gpt5_mini, .gpt5_nano, .gpt5_chat, .gpt4_1, .gpt4_o, .chatgpt_4o_latest,
+            .gpt5, .gpt5_mini, .gpt5_nano, .gpt5_chat, .gpt5_1, .gpt5_1_chat_latest, .gpt4_1, .gpt4_o, .chatgpt_4o_latest,
             // cost-optimized
             .gpt4_1_mini, .gpt4_1_nano, .gpt4_o_mini,
             .gpt4_turbo, .gpt4, .gpt3_5Turbo,

Sources/OpenAI/Public/Schemas/Generated/Components.swift

@@ -4861,13 +4861,14 @@ public enum Components {
         ///
         /// Constrains effort on reasoning for
         /// [reasoning models](https://platform.openai.com/docs/guides/reasoning).
-        /// Currently supported values are `minimal`, `low`, `medium`, and `high`. Reducing
+        /// Currently supported values are `none`, `minimal`, `low`, `medium`, and `high`. Reducing
         /// reasoning effort can result in faster responses and fewer tokens used
         /// on reasoning in a response.
         ///
         ///
         /// - Remark: Generated from `#/components/schemas/ReasoningEffort`.
         @frozen public enum ReasoningEffort: String, Codable, Hashable, Sendable, CaseIterable {
+            case none = "none"
             case minimal = "minimal"
             case low = "low"
             case medium = "medium"

Tests/OpenAITests/OpenAITestsDecoder.swift

@@ -418,6 +418,43 @@ class OpenAITestsDecoder: XCTestCase {
         XCTAssertNil(decodedNil.verbosity)
     }
 
+    func testChatQueryWithReasoningEffortNone() throws {
+        let chatQuery = ChatQuery(
+            messages: [
+                .init(role: .user, content: "Who are you?")!
+            ],
+            model: .gpt5_1,
+            reasoningEffort: ChatQuery.ReasoningEffort.none
+        )
+        let expectedValue = """
+            {
+                "model": "gpt-5.1",
+                "messages": [
+                    {
+                        "role": "user",
+                        "content": "Who are you?"
+                    }
+                ],
+                "reasoning_effort": "none",
+                "stream": false
+            }
+            """
+
+        let chatQueryAsDict = try jsonDataAsNSDictionary(JSONEncoder().encode(chatQuery))
+        let expectedValueAsDict = try jsonDataAsNSDictionary(expectedValue.data(using: .utf8)!)
+
+        XCTAssertEqual(chatQueryAsDict, expectedValueAsDict)
+    }
+
+    func testReasoningEffortDecodingNone() throws {
+        let json = """
+        { "effort": "none" }
+        """
+        let data = json.data(using: .utf8)!
+        let decoded = try JSONDecoder().decode(Components.Schemas.Reasoning.self, from: data)
+        XCTAssertEqual(decoded.effort, Components.Schemas.ReasoningEffort.none)
+    }
+
     func testEmbeddings() async throws {
         let data = """
         {