Fix Qwen3.5 tool calling

Libraries/MLXLMCommon/Tool/Parsers/XMLFunctionParser.swift

@@ -5,10 +5,13 @@ import Foundation
 /// Parser for XML function format: <function=name><parameter=key>value</parameter></function>
 /// Reference: https://github.com/ml-explore/mlx-lm/blob/main/mlx_lm/tool_parsers/qwen3_coder.py
 public struct XMLFunctionParser: ToolCallParser, Sendable {
-    public let startTag: String? = nil  // Inline format - no wrapper tags
-    public let endTag: String? = nil
+    public let startTag: String?
+    public let endTag: String?
 
-    public init() {}
+    public init(startTag: String, endTag: String) {
+        self.startTag = startTag
+        self.endTag = endTag
+    }
 
     public func parse(content: String, tools: [[String: any Sendable]]?) -> ToolCall? {
         // Pattern: <function=(content)</function> — [\s\S] matches newlines

Libraries/MLXLMCommon/Tool/ToolCallFormat.swift

@@ -70,8 +70,8 @@ public enum ToolCallFormat: String, Sendable, Codable, CaseIterable {
     /// Example: `<|tool_call_start|>[func(arg='value')]<|tool_call_end|>`
     case lfm2
 
-    /// XML function format used by Qwen3 Coder.
-    /// Example: `<function=name><parameter=key>value</parameter></function>`
+    /// XML function format used by Nemotron, Qwen3 Coder, Qwen3.5, and similar models.
+    /// Example: `<tool_call><function=name><parameter=key>value</parameter></function></tool_call>`
     case xmlFunction = "xml_function"
 
     /// GLM4 format with arg_key/arg_value tags.
@@ -106,7 +106,7 @@ public enum ToolCallFormat: String, Sendable, Codable, CaseIterable {
             return PythonicToolCallParser(
                 startTag: "<|tool_call_start|>", endTag: "<|tool_call_end|>")
         case .xmlFunction:
-            return XMLFunctionParser()
+            return XMLFunctionParser(startTag: "<tool_call>", endTag: "</tool_call>")
         case .glm4:
             return GLM4ToolCallParser()
         case .gemma:
@@ -145,6 +145,16 @@ public enum ToolCallFormat: String, Sendable, Codable, CaseIterable {
             return .gemma
         }
 
+        // Nemotron family (nemotron_h, etc.)
+        if type.hasPrefix("nemotron") {
+            return .xmlFunction
+        }
+
+        // Qwen3.5 family (qwen3_5, qwen3_5_moe, etc.)
+        if type.hasPrefix("qwen3_5") {
+            return .xmlFunction
+        }
+
         // Mistral3 family (mistral3, mistral3_text, etc.)
         if type.hasPrefix("mistral3") {
             return .mistral

Tests/MLXLMIntegrationTests/IntegrationTestModels.swift

@@ -8,14 +8,28 @@ import MLXVLM
 enum IntegrationTestModelIDs {
     static let llmModelId = "mlx-community/Qwen3-4B-Instruct-2507-4bit"
     static let vlmModelId = "mlx-community/Qwen3-VL-4B-Instruct-4bit"
+
+    static let lfm2ModelId = "mlx-community/LFM2-2.6B-Exp-4bit"
+    static let glm4ModelId = "mlx-community/GLM-4-9B-0414-4bit"
+    static let mistral3ModelId = "mlx-community/Ministral-3-3B-Instruct-2512-4bit"
+    static let nemotronModelId = "mlx-community/NVIDIA-Nemotron-3-Nano-30B-A3B-4bit"
+    static let qwen35ModelId = "mlx-community/Qwen3.5-2B-4bit"
 }
 
 actor IntegrationTestModels {
     static let shared = IntegrationTestModels()
 
+    private init() {}
+
     private var llmTask: Task<ModelContainer, Error>?
     private var vlmTask: Task<ModelContainer, Error>?
 
+    private var lfm2Task: Task<ModelContainer, Error>?
+    private var glm4Task: Task<ModelContainer, Error>?
+    private var mistral3Task: Task<ModelContainer, Error>?
+    private var nemotronTask: Task<ModelContainer, Error>?
+    private var qwen35Task: Task<ModelContainer, Error>?
+
     func llmContainer() async throws -> ModelContainer {
         if let task = llmTask {
             return try await task.value
@@ -43,4 +57,74 @@ actor IntegrationTestModels {
         vlmTask = task
         return try await task.value
     }
+
+    func lfm2Container() async throws -> ModelContainer {
+        if let task = lfm2Task {
+            return try await task.value
+        }
+
+        let task = Task {
+            try await LLMModelFactory.shared.loadContainer(
+                configuration: .init(id: IntegrationTestModelIDs.lfm2ModelId)
+            )
+        }
+        lfm2Task = task
+        return try await task.value
+    }
+
+    func glm4Container() async throws -> ModelContainer {
+        if let task = glm4Task {
+            return try await task.value
+        }
+
+        let task = Task {
+            try await LLMModelFactory.shared.loadContainer(
+                configuration: .init(id: IntegrationTestModelIDs.glm4ModelId)
+            )
+        }
+        glm4Task = task
+        return try await task.value
+    }
+
+    func mistral3Container() async throws -> ModelContainer {
+        if let task = mistral3Task {
+            return try await task.value
+        }
+
+        let task = Task {
+            try await LLMModelFactory.shared.loadContainer(
+                configuration: .init(id: IntegrationTestModelIDs.mistral3ModelId)
+            )
+        }
+        mistral3Task = task
+        return try await task.value
+    }
+
+    func nemotronContainer() async throws -> ModelContainer {
+        if let task = nemotronTask {
+            return try await task.value
+        }
+
+        let task = Task {
+            try await LLMModelFactory.shared.loadContainer(
+                configuration: .init(id: IntegrationTestModelIDs.nemotronModelId)
+            )
+        }
+        nemotronTask = task
+        return try await task.value
+    }
+
+    func qwen35Container() async throws -> ModelContainer {
+        if let task = qwen35Task {
+            return try await task.value
+        }
+
+        let task = Task {
+            try await LLMModelFactory.shared.loadContainer(
+                configuration: .init(id: IntegrationTestModelIDs.qwen35ModelId)
+            )
+        }
+        qwen35Task = task
+        return try await task.value
+    }
 }