@@ -42,11 +42,13 @@ function LLMScreen() {
   const { bottom } = useSafeAreaInsets();
   const [isTextInputFocused, setIsTextInputFocused] = useState(false);
   const [userInput, setUserInput] = useState('');
+  const [selectedModel, setSelectedModel] = useState<LLMModelSources>(
+    models.llm.lfm2_5_1_2b_instruct()
+  );
   const textInputRef = useRef<TextInput>(null);
   const { setGlobalGenerating } = useContext(GeneratingContext);
 
-  let llm = Object();
-  
+  const llm = useLLM({ model: selectedModel });
   const tokenCount = llm.isReady ? llm.getGeneratedTokenCount() : 0;
   const { stats, onMessageSend } = useLLMStats(
     llm.response,
@@ -110,6 +112,13 @@ function LLMScreen() {
               />
             </View>
           )}
+
+          <ModelPicker
+            models={LLM_MODELS}
+            selectedModel={selectedModel}
+            onSelect={(m) => setSelectedModel(m)}
+            disabled={llm.isGenerating}
+          />
           <StatsBar stats={stats} />
           <View
             style={[
 
@@ -3,7 +3,6 @@ import {
   LLAMA3_2_1B_QLORA,
   LLAMA3_2_3B_QLORA,
   LLMProps,
-  GEMMA4_E2B_QUANTIZED,
 } from 'react-native-executorch';
 import { ModelOption } from './ModelPicker';
 const llm = models.llm;
@@ -12,7 +11,7 @@ export type LLMModelSources = LLMProps['model'];
 
 export const LLM_MODELS: ModelOption<LLMModelSources>[] = [
   // Gemma4
-  { label: 'Gemma4 e2b Quantized', value: GEMMA4_E2B_QUANTIZED },
+  { label: 'Gemma4 E2B', value: llm.gemma4_e2b() },
   // Llama 3.2
   {
     label: 'Llama 3.2 1B',
 
@@ -17,6 +17,7 @@
 
 #include <rnexecutorch/metaprogramming/TypeConcepts.h>
 #include <rnexecutorch/models/instance_segmentation/Types.h>
+#include <rnexecutorch/models/llm/Types.h>
 #include <rnexecutorch/models/object_detection/Constants.h>
 #include <rnexecutorch/models/object_detection/Types.h>
 #include <rnexecutorch/models/ocr/Types.h>
@@ -318,6 +319,31 @@ getValue<std::span<uint64_t>>(const jsi::Value &val, jsi::Runtime &runtime) {
   return getTypedArrayAsSpan<uint64_t>(val, runtime);
 }
 
+template <typename T>
+inline void assignIfPresent(T &field, const jsi::Object &obj,
+                            jsi::Runtime &runtime, const char *name) {
+  if (!obj.hasProperty(runtime, name))
+    return;
+  jsi::Value v = obj.getProperty(runtime, name);
+  if (v.isUndefined() || v.isNull())
+    return;
+  field = getValue<T>(v, runtime);
+}
+
+template <>
+inline models::llm::MultimodalInputs
+getValue<models::llm::MultimodalInputs>(const jsi::Value &val,
+                                        jsi::Runtime &runtime) {
+  models::llm::MultimodalInputs multimodalInputs;
+  jsi::Object obj = val.asObject(runtime);
+  assignIfPresent(multimodalInputs.imagePaths, obj, runtime, "imagePaths");
+  assignIfPresent(multimodalInputs.imageToken, obj, runtime, "imageToken");
+  assignIfPresent(multimodalInputs.audioWaveforms, obj, runtime,
+                  "audioWaveforms");
+  assignIfPresent(multimodalInputs.audioToken, obj, runtime, "audioToken");
+  return multimodalInputs;
+}
+
 // Conversion from C++ types to jsi --------------------------------------------
 
 // Implementation functions might return any type, but in a promise we can only
 
@@ -76,10 +76,9 @@ std::string LLM::generate(std::string input,
   return output;
 }
 
-std::string LLM::generateMultimodal(
-    std::string prompt, std::shared_ptr<jsi::Function> callback,
-    std::vector<std::string> imagePaths, std::string imageToken,
-    std::vector<std::vector<float>> audioWaveforms, std::string audioToken) {
+std::string LLM::generateMultimodal(std::string prompt,
+                                    std::shared_ptr<jsi::Function> callback,
+                                    MultimodalInputs mutlimodalInputs) {
   if (!runner_ || !runner_->is_loaded()) {
     throw RnExecutorchError(RnExecutorchErrorCode::ModuleNotLoaded,
                             "Runner is not loaded");
@@ -88,7 +87,8 @@ std::string LLM::generateMultimodal(
     throw RnExecutorchError(RnExecutorchErrorCode::InvalidUserInput,
                             "This model does not support multimodal input.");
   }
-  if (imageToken.empty() && audioToken.empty()) {
+  if (mutlimodalInputs.imageToken.empty() &&
+      mutlimodalInputs.audioToken.empty()) {
     throw RnExecutorchError(
         RnExecutorchErrorCode::InvalidUserInput,
         "At least one of imageToken/audioToken must be non-empty");
@@ -98,12 +98,13 @@ std::string LLM::generateMultimodal(
   // so that image/audio placeholders can be freely interleaved in the prompt.
   std::vector<llm::MultimodalInput> inputs;
   size_t imageIdx = 0, audioIdx = 0, pos = 0;
-  constexpr int32_t kAudioSampleRate = 16000;
   while (pos < prompt.size()) {
-    size_t imgAt =
-        imageToken.empty() ? std::string::npos : prompt.find(imageToken, pos);
-    size_t audAt =
-        audioToken.empty() ? std::string::npos : prompt.find(audioToken, pos);
+    size_t imgAt = mutlimodalInputs.imageToken.empty()
+                       ? std::string::npos
+                       : prompt.find(mutlimodalInputs.imageToken, pos);
+    size_t audAt = mutlimodalInputs.audioToken.empty()
+                       ? std::string::npos
+                       : prompt.find(mutlimodalInputs.audioToken, pos);
     if (imgAt == std::string::npos && audAt == std::string::npos) {
       inputs.push_back(llm::make_text_input(prompt.substr(pos)));
       break;
@@ -115,25 +116,27 @@ std::string LLM::generateMultimodal(
       inputs.push_back(llm::make_text_input(prompt.substr(pos, at - pos)));
     }
     if (imageFirst) {
-      if (imageIdx >= imagePaths.size()) {
+      if (imageIdx >= mutlimodalInputs.imagePaths.size()) {
         throw RnExecutorchError(RnExecutorchErrorCode::InvalidUserInput,
-                                "More '" + imageToken +
+                                "More '" + mutlimodalInputs.imageToken +
                                     "' placeholders than image paths");
       }
-      inputs.push_back(llm::make_image_input(imagePaths[imageIdx++]));
-      pos = at + imageToken.size();
+      inputs.push_back(
+          llm::make_image_input(mutlimodalInputs.imagePaths[imageIdx++]));
+      pos = at + mutlimodalInputs.imageToken.size();
     } else {
-      if (audioIdx >= audioWaveforms.size()) {
+      if (audioIdx >= mutlimodalInputs.audioWaveforms.size()) {
         throw RnExecutorchError(RnExecutorchErrorCode::InvalidUserInput,
-                                "More '" + audioToken +
+                                "More '" + mutlimodalInputs.audioToken +
                                     "' placeholders than audio waveforms");
       }
       inputs.push_back(llm::make_audio_input(
-          std::move(audioWaveforms[audioIdx++]), kAudioSampleRate));
-      pos = at + audioToken.size();
+          std::move(mutlimodalInputs.audioWaveforms[audioIdx++])));
+      pos = at + mutlimodalInputs.audioToken.size();
     }
   }
-  if (imageIdx < imagePaths.size() || audioIdx < audioWaveforms.size()) {
+  if (imageIdx < mutlimodalInputs.imagePaths.size() ||
+      audioIdx < mutlimodalInputs.audioWaveforms.size()) {
     throw RnExecutorchError(
         RnExecutorchErrorCode::InvalidUserInput,
         "More image/audio paths provided than placeholders in prompt");
 
@@ -7,6 +7,7 @@
 #include <ReactCommon/CallInvoker.h>
 #include <jsi/jsi.h>
 #include <rnexecutorch/models/BaseModel.h>
+#include <rnexecutorch/models/llm/Types.h>
 #include <runner/base_llm_runner.h>
 
 namespace rnexecutorch {
@@ -27,11 +28,9 @@ class LLM : public BaseModel {
   // The prompt is scanned for `imageToken` and/or `audioToken` placeholders;
   // each placeholder consumes the next entry from its respective vector in
   // order. Either set of paths/waveforms/token may be empty.
-  std::string generateMultimodal(
-      std::string prompt, std::shared_ptr<jsi::Function> callback,
-      std::vector<std::string> imagePaths = {}, std::string imageToken = "",
-      std::vector<std::vector<float>> audioWaveforms = {},
-      std::string audioToken = "");
+  std::string generateMultimodal(std::string prompt,
+                                 std::shared_ptr<jsi::Function> callback,
+                                 MultimodalInputs mutlimodalInputs = {});
 
   void interrupt();
   void reset();
 
@@ -0,0 +1,15 @@
+#pragma once
+
+#include <string>
+#include <vector>
+
+namespace rnexecutorch::models::llm {
+struct MultimodalInputs {
+  std::vector<std::string> imagePaths;
+  std::string imageToken;
+  std::vector<std::vector<float>> audioWaveforms;
+  std::string audioToken;
+};
+using Token = uint64_t;
+
+} // namespace rnexecutorch::models::llm
@@ -18,8 +18,8 @@ BaseLLMRunner::BaseLLMRunner(std::unique_ptr<Module> module,
       tokenizer_(std::make_unique<tokenizers::HFTokenizer>()),
       metadata_({
           {kEnableDynamicShape, false},
-          {kMaxSeqLen, 2048},
-          {kMaxContextLen, 2048},
+          {kMaxSeqLen, 128},
+          {kMaxContextLen, 128},
           {kUseKVCache, true},
       }) {}
 
 
@@ -2,7 +2,6 @@
 #include "audio_encoder.h"
 
 #include <rnexecutorch/Error.h>
-#include <rnexecutorch/Log.h>
 #include <runner/constants.h>
 
 #include <executorch/extension/tensor/tensor.h>
@@ -21,12 +20,7 @@ using ::executorch::runtime::EValue;
 using ::executorch::runtime::Result;
 
 namespace {
-// Matches AUDIO_SAMPLES_PER_BLOCK in gemma_export/experiments_vulkan/
-// op_bisect/iter201_mm_4method_dynaudio_prefill2048_export.py.
-// The PTE's audio_samples dim was exported as `7680 * audio_blocks`.
 constexpr int32_t kSamplesPerBlock = 7680;
-// k ∈ [kAudioBlockKMin, kAudioBlockKMax] from MODEL_INTERFACE.md §6.
-// k=62 == 29.76 s @ 16 kHz is the SDPA mask + rel-shift bake point.
 constexpr int64_t kAudioBlockKMin = 1;
 constexpr int64_t kAudioBlockKMax = 62;
 } // namespace
@@ -55,7 +49,9 @@ bool AudioEncoder::is_loaded() const noexcept {
   return module_->is_method_loaded(kAudioEncoderMethod);
 }
 
-int32_t AudioEncoder::encoderTokenCount() const { return last_token_count_; }
+int32_t AudioEncoder::encoderTokenCount() const noexcept {
+  return last_token_count_;
+}
 
 Result<EValue> AudioEncoder::encode(const MultimodalInput &input) {
   if (!is_loaded()) {
@@ -68,9 +64,6 @@ Result<EValue> AudioEncoder::encode(const MultimodalInput &input) {
   const auto &wav = input.get_audio();
   ET_CHECK_OR_RETURN_ERROR(!wav.samples.empty(), InvalidArgument,
                            "AudioEncoder: empty waveform");
-  ET_CHECK_OR_RETURN_ERROR(
-      wav.sample_rate == 16000, InvalidArgument,
-      "AudioEncoder: expected 16000 Hz waveform, got %d Hz", wav.sample_rate);
 
   const int64_t n_valid = static_cast<int64_t>(wav.samples.size());
   const int64_t k_blocks = (n_valid + kSamplesPerBlock - 1) / kSamplesPerBlock;
@@ -116,10 +109,6 @@ Result<EValue> AudioEncoder::encode(const MultimodalInput &input) {
                            "audio_encoder output rank=%zd, expected 3",
                            audio_tensor.dim());
   last_token_count_ = static_cast<int32_t>(audio_tensor.size(1));
-  rnexecutorch::log(rnexecutorch::LOG_LEVEL::Info,
-                    "AudioEncoder: valid_samples=", n_valid,
-                    " padded_samples=", n_padded, " k_blocks=", k_blocks,
-                    " audio_tokens=", last_token_count_);
   return exec_result[0];
 }
 
 
@@ -28,7 +28,7 @@ class AudioEncoder : public IEncoder {
   encode(const MultimodalInput &input) override;
   // Number of audio embedding tokens produced per encode() call. 0 until first
   // encode, since Gemma4's audio_encoder has a dynamic T dim.
-  int32_t encoderTokenCount() const override;
+  int32_t encoderTokenCount() const noexcept override;
 
 private:
   ::executorch::extension::Module *module_;