Fix downloads: HuggingFace auth + Gemma 3 1B IT models

Google moved all models to HuggingFace (old CDN returns 404).
All downloads now require a HuggingFace API token.

- Add HF token input field (saved to SharedPreferences, restored on relaunch)
- Download uses Bearer auth header
- Switch to Gemma 3 1B IT (litert-community) — .task format, works with
  current MediaPipe API, 555 MB Q4 or 1 GB Q8
- Clear auth error messages (401/403 shown to user)
- Gemma 3n E4B/E2B (.litertlm format) requires runtime upgrade — planned next

Co-Authored-By: Claude Sonnet 4.6 <[email protected]>

app/src/main/java/com/pixel10/ai/inference/ModelDownloader.kt

@@ -10,23 +10,44 @@ import java.net.HttpURLConnection
 import java.net.URL
 
 /**
- * Downloads a MediaPipe-compatible Gemma model for background-safe inference.
+ * Downloads a MediaPipe-compatible model for background-safe inference.
  *
- * Gemini Nano (ML Kit) blocks inference when the app is backgrounded (ErrorCode 30).
- * MediaPipe with a local model file has no such restriction — it runs entirely in
- * the app process using the Tensor G5 GPU via OpenCL/Vulkan.
+ * All models are hosted on HuggingFace and require a free API token.
+ * Get one at: https://huggingface.co/settings/tokens
  *
- * The downloaded model is stored in the app's private files directory and
- * survives app restarts. Only needs to be downloaded once (~1.3 GB).
+ * Models use the MediaPipe `.task` format, compatible with [MediaPipeModel].
+ * Gemma 3n E4B/E2B (`.litertlm` format) requires a runtime upgrade — coming later.
  */
 object ModelDownloader {
 
     private const val TAG = "ModelDownloader"
+    private const val HF_BASE = "https://huggingface.co"
 
-    const val MODEL_FILENAME = "gemma-2b-it-gpu-int4.bin"
-    private const val MODEL_URL =
-        "https://storage.googleapis.com/mediapipe-models/llm_inference/" +
-        "gemma-2b-it-gpu-int4/float16/1/gemma-2b-it-gpu-int4.bin"
+    /** Available model specs downloadable from HuggingFace. */
+    enum class ModelSpec(
+        val displayName: String,
+        val filename: String,
+        val repo: String,
+        val sizeMb: Int,
+        val description: String
+    ) {
+        /** Recommended: best size/quality trade-off, runs fast on Tensor G5. */
+        GEMMA_3_1B_Q4(
+            displayName = "Gemma 3 1B IT (Q4)",
+            filename = "gemma3-1b-it-int4.task",
+            repo = "litert-community/Gemma3-1B-IT",
+            sizeMb = 555,
+            description = "Best balance — fast & capable (~555 MB)"
+        ),
+        /** Higher quality, slower. Good for complex reasoning. */
+        GEMMA_3_1B_Q8(
+            displayName = "Gemma 3 1B IT (Q8)",
+            filename = "gemma3-1b-it-int8-web.task",
+            repo = "litert-community/Gemma3-1B-IT",
+            sizeMb = 1010,
+            description = "Higher quality, slower (~1 GB)"
+        )
+    }
 
     data class Progress(
         val downloadedBytes: Long,
@@ -34,34 +55,58 @@ object ModelDownloader {
         val percent: Int = if (totalBytes > 0) (downloadedBytes * 100 / totalBytes).toInt() else 0
     )
 
+    /** Returns true if any supported model is present in the app's files directory. */
     fun isModelPresent(context: Context): Boolean =
-        modelFile(context).let { it.exists() && it.length() > 1_000_000L }
+        ModelSpec.values().any { modelFile(context, it).let { f -> f.exists() && f.length() > 1_000_000L } }
 
-    fun modelFile(context: Context): File = File(context.filesDir, MODEL_FILENAME)
+    /** Returns the installed [ModelSpec], or null if no model is present. */
+    fun installedSpec(context: Context): ModelSpec? =
+        ModelSpec.values().firstOrNull { modelFile(context, it).let { f -> f.exists() && f.length() > 1_000_000L } }
+
+    fun modelFile(context: Context, spec: ModelSpec): File =
+        File(context.filesDir, spec.filename)
+
+    /** Legacy compat — returns the file of the installed model, or Q4 path as default. */
+    fun modelFile(context: Context): File =
+        installedSpec(context)?.let { modelFile(context, it) }
+            ?: modelFile(context, ModelSpec.GEMMA_3_1B_Q4)
 
     /**
-     * Download the model, reporting progress via [onProgress].
+     * Download [spec] from HuggingFace, using [hfToken] for authentication.
      * Supports resume — if a partial file exists, continues from where it left off.
+     *
+     * Get a free token at https://huggingface.co/settings/tokens
      */
     suspend fun download(
         context: Context,
+        spec: ModelSpec = ModelSpec.GEMMA_3_1B_Q4,
+        hfToken: String,
         onProgress: (Progress) -> Unit
     ) = withContext(Dispatchers.IO) {
-        val dest = modelFile(context)
+        if (hfToken.isBlank()) throw OnDeviceModel.InferenceException(
+            "HuggingFace token required.\nGet a free token at huggingface.co/settings/tokens"
+        )
+
+        val dest = modelFile(context, spec)
         val alreadyDownloaded = if (dest.exists()) dest.length() else 0L
+        val downloadUrl = "$HF_BASE/${spec.repo}/resolve/main/${spec.filename}"
 
-        Log.i(TAG, "Download starting (already have $alreadyDownloaded bytes)")
+        Log.i(TAG, "Download starting ${spec.displayName} from $downloadUrl (already have $alreadyDownloaded bytes)")
 
-        val conn = URL(MODEL_URL).openConnection() as HttpURLConnection
+        val conn = URL(downloadUrl).openConnection() as HttpURLConnection
         try {
             conn.connectTimeout = 30_000
             conn.readTimeout   = 60_000
+            conn.setRequestProperty("Authorization", "Bearer $hfToken")
             if (alreadyDownloaded > 0) {
                 conn.setRequestProperty("Range", "bytes=$alreadyDownloaded-")
             }
             conn.connect()
 
             val code = conn.responseCode
+            if (code == 401 || code == 403) throw OnDeviceModel.InferenceException(
+                "Authentication failed (HTTP $code).\nCheck your HuggingFace token."
+            )
             val resuming = code == HttpURLConnection.HTTP_PARTIAL   // 206
             if (code != HttpURLConnection.HTTP_OK && !resuming) {
                 throw OnDeviceModel.InferenceException("Download failed: HTTP $code")
@@ -90,7 +135,7 @@ object ModelDownloader {
     }
 
     fun deleteModel(context: Context) {
-        modelFile(context).delete()
-        Log.i(TAG, "Model deleted")
+        ModelSpec.values().forEach { modelFile(context, it).delete() }
+        Log.i(TAG, "All models deleted")
     }
 }

app/src/main/java/com/pixel10/ai/ui/MainActivity.kt

@@ -5,6 +5,7 @@ import android.content.ComponentName
 import android.content.Context
 import android.content.Intent
 import android.content.ServiceConnection
+import android.content.SharedPreferences
 import android.content.pm.PackageManager
 import android.graphics.drawable.GradientDrawable
 import android.net.wifi.WifiManager
@@ -19,6 +20,7 @@ import androidx.lifecycle.lifecycleScope
 import com.pixel10.ai.R
 import com.pixel10.ai.databinding.ActivityMainBinding
 import com.pixel10.ai.inference.ModelDownloader
+import com.pixel10.ai.inference.ModelDownloader.ModelSpec
 import com.pixel10.ai.server.ApiServerService
 import kotlinx.coroutines.launch
 import java.text.SimpleDateFormat
@@ -28,6 +30,7 @@ import java.util.Locale
 class MainActivity : AppCompatActivity() {
 
     private lateinit var binding: ActivityMainBinding
+    private lateinit var prefs: SharedPreferences
     private var service: ApiServerService? = null
     private var bound = false
     private var downloading = false
@@ -67,14 +70,23 @@ class MainActivity : AppCompatActivity() {
         binding = ActivityMainBinding.inflate(layoutInflater)
         setContentView(binding.root)
 
+        prefs = getSharedPreferences("pixel10_prefs", MODE_PRIVATE)
         requestNotificationPermission()
 
+        // Restore saved HF token
+        binding.etHfToken.setText(prefs.getString("hf_token", ""))
+
         binding.btnToggle.setOnClickListener {
             if (service?.isRunning == true) stopServer() else startServer()
         }
 
         binding.btnDownloadModel.setOnClickListener {
-            startModelDownload()
+            saveHfToken()
+            startModelDownload(ModelSpec.GEMMA_3_1B_Q4)
+        }
+        binding.btnDownloadGemma3Q8.setOnClickListener {
+            saveHfToken()
+            startModelDownload(ModelSpec.GEMMA_3_1B_Q8)
         }
 
         updateModelCard()
@@ -121,34 +133,43 @@ class MainActivity : AppCompatActivity() {
         }
     }
 
-    private fun startModelDownload() {
+    private fun saveHfToken() {
+        val token = binding.etHfToken.text.toString().trim()
+        prefs.edit().putString("hf_token", token).apply()
+    }
+
+    private fun startModelDownload(spec: ModelSpec) {
         if (downloading) return
+        val token = binding.etHfToken.text.toString().trim()
+        if (token.isBlank()) {
+            binding.tvModelDownloadStatus.text = "Enter your HuggingFace token first"
+            return
+        }
         downloading = true
-        binding.btnDownloadModel.isEnabled = false
-        binding.btnDownloadModel.text = "Downloading…"
+        setDownloadButtonsEnabled(false)
         binding.progressDownload.visibility = View.VISIBLE
-        binding.tvModelDownloadStatus.text = "Starting download…"
+        binding.tvModelDownloadStatus.text = "Starting download: ${spec.displayName}…"
 
         lifecycleScope.launch {
             try {
-                ModelDownloader.download(this@MainActivity) { progress ->
+                ModelDownloader.download(this@MainActivity, spec, token) { progress ->
                     runOnUiThread {
                         binding.progressDownload.progress = progress.percent
                         val mb = progress.downloadedBytes / 1_048_576
                         val total = progress.totalBytes / 1_048_576
-                        binding.tvModelDownloadStatus.text = "Downloading… ${mb}MB / ${total}MB (${progress.percent}%)"
+                        binding.tvModelDownloadStatus.text =
+                            "${spec.displayName}: ${mb}MB / ${total}MB (${progress.percent}%)"
                     }
                 }
                 runOnUiThread {
                     downloading = false
                     updateModelCard()
-                    appendLog("Model downloaded — background inference enabled")
+                    appendLog("${spec.displayName} downloaded — background inference enabled")
                 }
             } catch (e: Exception) {
                 runOnUiThread {
                     downloading = false
-                    binding.btnDownloadModel.isEnabled = true
-                    binding.btnDownloadModel.text = getString(R.string.btn_download_model)
+                    setDownloadButtonsEnabled(true)
                     binding.progressDownload.visibility = View.GONE
                     binding.tvModelDownloadStatus.text = "Download failed: ${e.message}"
                     appendLog("Download error: ${e.message}")
@@ -157,17 +178,25 @@ class MainActivity : AppCompatActivity() {
         }
     }
 
+    private fun setDownloadButtonsEnabled(enabled: Boolean) {
+        binding.btnDownloadModel.isEnabled = enabled
+        binding.btnDownloadGemma3Q8.isEnabled = enabled
+    }
+
     private fun updateModelCard() {
-        val present = ModelDownloader.isModelPresent(this)
-        if (present) {
-            binding.tvModelDownloadStatus.text = getString(R.string.model_downloaded)
+        val spec = ModelDownloader.installedSpec(this)
+        if (spec != null) {
+            binding.tvModelDownloadStatus.text = getString(R.string.model_downloaded, spec.displayName)
+            binding.etHfToken.visibility = View.GONE
             binding.btnDownloadModel.visibility = View.GONE
+            binding.btnDownloadGemma3Q8.visibility = View.GONE
             binding.progressDownload.visibility = View.GONE
         } else {
             binding.tvModelDownloadStatus.text = getString(R.string.model_not_downloaded)
+            binding.etHfToken.visibility = View.VISIBLE
             binding.btnDownloadModel.visibility = View.VISIBLE
-            binding.btnDownloadModel.isEnabled = true
-            binding.btnDownloadModel.text = getString(R.string.btn_download_model)
+            binding.btnDownloadGemma3Q8.visibility = View.VISIBLE
+            setDownloadButtonsEnabled(true)
             binding.progressDownload.visibility = View.GONE
         }
     }

app/src/main/res/layout/activity_main.xml

@@ -132,6 +132,19 @@
                 android:textColor="@color/log_text"
                 android:textSize="13sp" />
 
+            <com.google.android.material.textfield.TextInputEditText
+                android:id="@+id/etHfToken"
+                android:layout_width="match_parent"
+                android:layout_height="48dp"
+                android:layout_marginTop="8dp"
+                android:hint="@string/hf_token_hint"
+                android:inputType="textPassword"
+                android:textColor="@color/on_surface"
+                android:textColorHint="@color/log_text"
+                android:textSize="13sp"
+                android:fontFamily="monospace"
+                android:backgroundTint="@color/primary" />
+
             <ProgressBar
                 android:id="@+id/progressDownload"
                 style="@android:style/Widget.ProgressBar.Horizontal"
@@ -147,7 +160,17 @@
                 android:layout_width="wrap_content"
                 android:layout_height="wrap_content"
                 android:layout_marginTop="8dp"
-                android:text="@string/btn_download_model"
+                android:text="@string/btn_download_gemma3_q4"
+                android:textSize="13sp"
+                app:cornerRadius="8dp" />
+
+            <com.google.android.material.button.MaterialButton
+                android:id="@+id/btnDownloadGemma3Q8"
+                style="@style/Widget.MaterialComponents.Button.OutlinedButton"
+                android:layout_width="wrap_content"
+                android:layout_height="wrap_content"
+                android:layout_marginTop="4dp"
+                android:text="@string/btn_download_gemma3_q8"
                 android:textSize="13sp"
                 app:cornerRadius="8dp" />
         </LinearLayout>

app/src/main/res/values/strings.xml

@@ -18,9 +18,11 @@
     <string name="model_not_loaded">Model: not loaded</string>
 
     <!-- Controls -->
-    <string name="btn_download_model">Download Model (~1.3 GB)</string>
-    <string name="model_downloaded">✓ Gemma 2B ready — background inference enabled</string>
-    <string name="model_not_downloaded">No local model. Download to enable background inference.</string>
+    <string name="hf_token_hint">HuggingFace token (huggingface.co/settings/tokens)</string>
+    <string name="btn_download_gemma3_q4">⭐ Gemma 3 1B IT Q4 — Fast (~555 MB)</string>
+    <string name="btn_download_gemma3_q8">Gemma 3 1B IT Q8 — Higher quality (~1 GB)</string>
+    <string name="model_downloaded">✓ %s ready — background inference enabled</string>
+    <string name="model_not_downloaded">No local model. Enter HuggingFace token and download.</string>
     <string name="btn_start">Start Server</string>
     <string name="btn_stop">Stop Server</string>
     <string name="port_label">Port:</string>