From ebc113ee07d3a39195d49f59e495395edbe40138 Mon Sep 17 00:00:00 2001
From: Theo Weijmer <t.weijmer@pensare.nl>
Date: Thu, 1 Oct 2026 09:58:09 +0200
Subject: [PATCH] fix: store model in Application Support, load at startup,
 skip network when cached

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
---
 .../SstWhisper/WhisperKitTranscriber.swift    | 59 ++++++++++++++++++-
 1 file changed, 57 insertions(+), 2 deletions(-)

diff --git a/Sources/SstWhisper/WhisperKitTranscriber.swift b/Sources/SstWhisper/WhisperKitTranscriber.swift
index fe8a520..6bf4916 100644
--- a/Sources/SstWhisper/WhisperKitTranscriber.swift
+++ b/Sources/SstWhisper/WhisperKitTranscriber.swift
@@ -12,13 +12,68 @@ public final class WhisperKitTranscriber: Transcriber {
     private let modelName: String
     private var pipeline: WhisperKit?
 
+    /// Download progress (0...1) during the one-time model download. May be called from any thread.
+    public var onProgress: ((Double) -> Void)?
+
     public init(model: String = "large-v3-v20240930_turbo") {
         self.modelName = model
     }
 
-    /// Downloads the model on first run (≈1.5 GB) and loads it into memory.
+    /// Hugging Face hub root; WhisperKit stores models in `<root>/models/<repo>/<variant>`.
+    /// Kept in Application Support so it is neither in ~/Documents (TCC prompt, iCloud) nor re-fetched.
+    private static var storageBase: URL {
+        FileManager.default.urls(for: .applicationSupportDirectory, in: .userDomainMask)[0]
+            .appending(path: "Sst/models", directoryHint: .isDirectory)
+    }
+
+    private static let repo = "argmaxinc/whisperkit-coreml"
+
+    /// A previously downloaded variant folder, if it contains the compiled Core ML models.
+    private func localModelFolder(base: URL) -> URL? {
+        let repoDir = base.appending(path: "models/\(Self.repo)", directoryHint: .isDirectory)
+        let folders = (try? FileManager.default.contentsOfDirectory(atPath: repoDir.path)) ?? []
+        return folders
+            .filter { $0.hasSuffix(modelName) }
+            .map { repoDir.appending(path: $0, directoryHint: .isDirectory) }
+            .first { folder in
+                ["MelSpectrogram", "AudioEncoder", "TextDecoder"].allSatisfy {
+                    FileManager.default.fileExists(
+                        atPath: folder.appending(path: "\($0).mlmodelc/coremldata.bin").path)
+                }
+            }
+    }
+
+    private func loadPipeline(from folder: URL, base: URL) async throws -> WhisperKit {
+        try await WhisperKit(WhisperKitConfig(
+            downloadBase: base,
+            modelFolder: folder.path,
+            tokenizerFolder: base,
+            prewarm: true,
+            load: true,
+            download: false
+        ))
+    }
+
+    /// Loads the model into memory. Downloads it (≈1.5 GB) only if it is not stored locally yet;
+    /// once downloaded, launching needs no network.
     public func load() async throws {
-        pipeline = try await WhisperKit(WhisperKitConfig(model: modelName))
+        guard pipeline == nil else { return }
+        let base = Self.storageBase
+        try FileManager.default.createDirectory(at: base, withIntermediateDirectories: true)
+
+        if let folder = localModelFolder(base: base),
+           let loaded = try? await loadPipeline(from: folder, base: base) {
+            pipeline = loaded
+            return
+        }
+        let folder = try await WhisperKit.download(
+            variant: modelName,
+            downloadBase: base,
+            from: Self.repo
+        ) { [onProgress] progress in
+            onProgress?(progress.fractionCompleted)
+        }
+        pipeline = try await loadPipeline(from: folder, base: base)
     }
 
     public func transcribe(_ samples: [Float]) async throws -> String {