NewYour coding agent can read the release notes before it upgrades.Set up the MCP server β
pub.dev, #1807 most downloaded on pub.dev
Moved to flutter_edge_ai. This is the final release under the name flutter_gemma; new versions ship as flutter_edge_ai, part of Flutter Edge AI.
Latest stable
1.11.4
Last release 1 months ago
05 Oct 2026
Ships on a steady schedule
a new release about every 1 weeks
Nearly every release is documented
notes for 60 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
3 years old
142 releases Β· first in 2024
Drag across the ruler to compare, or tap a release to open it.
β‘ MTP / speculative decoding for Gemma 4 β LiteRT-LM 0.11.0 enables Multi-Token Prediction on macOS / iOS / Android / Windows. New enableSpeculativeDe
enableSpeculativeDecoding flag on getActiveModel() overrides the model default if needed..litertlm (regression fix from 0.14.0) β PreferredBackend.npu again drives Qualcomm QNN / Google Tensor / MediaTek dispatch libs through LiteRT-LM's Backend::NPU.PreferredBackend.npu accepted on macOS / Linux / Windows (#261). Same dispatch-lib requirement as Android.google-ai-edge/LiteRT-LM#2225. Workaround in README β Linux Setup..flutter_gemma_native_version) β bumps clean stale companion libs automatically.See CHANGELOG.md for full release notes.
enableSpeculativeDecoding flag on getActiveModel() (null = model default; true/false to override)..litertlm (regression from 0.14.0): PreferredBackend.npu on Android .litertlm models routes through LiteRT-LM's Backend::NPU again β same as 0.13.x's Kotlin path before the FFI migration silently dropped it. Requires a Qualcomm QNN / Google Tensor / MediaTek dispatch lib on the device; without one, engine_create fails with a dispatch error. MediaPipe .task models still don't support NPU (MediaPipe SDK limitation, unchanged).PreferredBackend.npu on desktop (#261): macOS / Linux / Windows backend arm wired. Same dispatch-lib requirement as Android.libLiteRtWebGpuAccelerator.so segfaults during graph compile (filed google-ai-edge/LiteRT-LM#2225) and multi-signature vision encoder rejected by native (engine_create error). Workaround: pin pre-MTP HF revision 7fa1d78473894f7e736a21d920c3aa80f950c0db for Linux users. Other platforms unaffected..flutter_gemma_native_version) β bumps clean stale companion libs automatically.Your coding agent can read these notes before it upgrades. Set up the MCP server β
0.14.5: restore desktop tensorflowlite_c bundling via Native Assets by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/258
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.14.4...v0.14.5
tensorflowlite_c.{dll,so,dylib} now bundled via Native Assets β regression from 0.14.0 setup-script removal.0.14.4: fix macOS dylib loading on pub.dev (#255) + native lib size reduction by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/256
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.14.3...v0.14.4
fromAsset install on desktop (#250 mode 2).0.14.3: iOS minOS patch (#245) + Android 16KB pages (#253) + web RAG pipeline (#251) by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.14.2...v0.14.3
libGemmaModelConstraintProvider.dylib minos 26.2 β 14.0 to match the other companion dylibs.libLiteRtLm.so with -Wl,-z,max-page-size=16384.npm run build, bundles cache_api.js + LiteRT WASM into dist/, ships package.json + vite.config.js in pub tarball.0.14.2: macOS install_name_tool fix (#247) + Gemma 4 escape tokens (#248) + ABI matrix by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pul
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.14.1...v0.14.2
install_name_tool failure (#247): dart run / build_runner / flutter test on a pure-Dart library aborted with larger updated load commands do not fit because upstream Apple companion dylibs lack -headerpad_max_install_names. Skip them from Native Assets on macOS and bundle via Podfile post_install instead.<|"|> tokens from tool_calls.arguments were written to history raw, making the model echo them on later turns. Strip recursively before persist via new SdkResponseParser.cleanRawForHistory.modelFromAsset install on desktop (#250): the large_file_handler channel call threw MissingPluginException on macOS / Windows / Linux because the package only ships Android + iOS plugins. Now catches it and falls back to the in-memory loadAsset β writeFile path. Embedding (localagents-rag) and .litertlm FFI paths additionally fail fast with a typed message on non-arm64 Android (x86_64 emulator / armeabi-v7a) instead of a generic JVM crash β see new "Platform & Architecture Support" section in README.macos/Podfile post_install to the new snippet (see README β macOS Setup). iOS / Linux / Windows / Android unaffected.[*/perf] lines break out cost of dylib load, engine_create, prefill, decode β visible via debugPrint in both debug and release builds.Trim 600+ KB of junk from the pub publish archive by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/241
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.14.0...v0.14.1
core/ffi/litert_lm_client.dart (which imports dart:ffi) through the flutter_gemma_interface.dart β mobile/flutter_gemma_mobile.dart chain. JS/Wasm targets cannot compile dart:ffi. Added conditional imports in mobile/flutter_gemma_mobile.dart that swap litert_lm_client.dart and ffi_inference_model.dart for *_stub.dart shims on dart.library.js_interop. The web plugin (FlutterGemmaWeb) registers itself as FlutterGemmaPlugin.instance before any FFI code path runs, so the stubs' constructors (which throw UnsupportedError) are never actually invoked on web.lib*.dylib symlinks alongside the bundled .framework/ accelerators in Runner.app/Frameworks/ so LiteRT-LM's gpu_registry could dlopen them by basename β App Store Connect rejected those builds with "Unexpected file found in Frameworks". 0.14.1 patches the upstream LiteRT-LM source (runtime/components/sampler_factory.cc and litert/runtime/accelerators/gpu_registry.cc) to load Apple platform accelerators via @executable_path/../Frameworks/<X>.framework/<X> (macOS) / @executable_path/Frameworks/<X>.framework/<X> (iOS) instead of libX.dylib. Native Assets bundles the framework bundles correctly out of the box; no host-side Podfile symlinks needed. Patch applied during local Bazel rebuild β see native/litert_lm/patch_c_api.sh section 10.ModelType.gemma4 routes tool definitions to LiteRT-LM SDK via litert_lm_conversation_config_set_tools (OpenAI Chat Completions JSON). SDK applies chat_template.jinja through minja, renders native <|tool>declaration:...<tool|> tokens, and parses the model's <|tool_call>...<tool_call|> response back into structured tool_calls JSON. flutter_gemma reads the result via SdkResponseParser.extractToolCalls (handles parallel calls and the multimodal content[] path) and returns FunctionCallResponse to the app β no Dart-side prompt engineering needed.<|"|> Gemma 4 escape tokens from string arguments (recursively, including nested maps/lists).example/lib/models/model.dart: Gemma 4 E2B / E4B entries switched to modelType: ModelType.gemma4.Release 0.13.3: fix macOS SIGSEGV, iOS pod install, Windows >2GB, iOS simulator build by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.13.3...v0.14.0
dart:ffi against the C API. Removed Kotlin/JVM gRPC server, Azul Zulu JRE 24 download, and litertlm-server.jar bundling. Engine creation ~2 s (was ~10β15 s incl. JVM cold-start).litertlm models on iPhone (Gemma 3 1B, Gemma 3n E2B, Gemma 4 E2B). Multimodal vision + audio work on devicedxil.dll + dxcompiler.dll v1.9.2602) bundled in the Windows native archive β no manual install required.litertlm models on Android now go through the same Dart FFI path as desktop/iOS (was com.google.ai.edge.litertlm:litertlm-android AAR before). MediaPipe stays for .task/.bin modelsLiteRtLmFfiClient (lib/core/ffi/)stream_proxy_redirect_stderr exposes glog/abseil output on iOS/Android via temp file; helps diagnose engine init failureshook/build.dart from GitHub release native-v0.10.2; SHA256-verified, bundled via Native Assetspost_install block in your Podfile to symlink lib*.dylib next to the bundled .frameworks β gpu_registry calls dlopen by basename. See the macOS Setup section in README for the exact snippet (iOS works via the same pattern in example/ios/Podfile)ModelType.qwen3: New model type for Qwen3 models with thinking support
/no_think appended automatically when isThinking: false β faster TTFTcreateChat(maxFunctionBufferLength: 2048) for long function call argsFileSource now handles backslash paths correctlyVectorStoreRepository.removeDocument(id:) to delete documents from vector storeFix Qwen3 thinking mode (#224): Qwen3 tags now stripped automatically
<think> tags now stripped automaticallyFix iOS compile error (#222): XNNPack delegate type mismatch in EmbeddingModel.swift
EmbeddingModel.swiftTensorFlowLiteSelectTfOps β simulator builds work on Apple SiliconFix macOS SIGSEGV (#219): Per-conversation mutex in gRPC server prevents conversation.close() racing with sendMessageAsync on a native thread β use-af
conversation.close() racing with sendMessageAsync on a native thread β use-after-free in C++ fixedsetup_desktop.sh now downloads libLiteRtMetalAccelerator.dylib from GitHub Release so GPU inference uses the Metal delegate instead of falling back to static C APITensorFlowLiteSwift (source pod β cloned entire TensorFlow repo) with direct TensorFlowLiteC C API in EmbeddingModel.swiftFileSource absolute paths: Accept both Unix (/path) and Windows (C:\path) absolute paths in FileSource validation
/path) and Windows (C:\path) absolute paths in FileSource validationLiteRT-LM 0.10.0: Updated Android and JVM SDK from 0.9.0 to 0.10.0
isThinking: true now works with Gemma 4 E2B/E4B models via extraContext and streaming thought channelslarge_file_handler platform support: Conditional imports for pub.dev platform analysis compatibilityisThinking: true now works with Gemma 4 E2B/E4B models (Android, iOS, Desktop; not Web)large_file_handler platform support: Conditional imports for pub.dev platform analysis compatibilityGemma 4 E2B/E4B: Added support for next-gen multimodal models (text + image + audio)
createChat() and createSession() for setting system-level context.litertlm models across platforms.litertlm models now work on iOS.litertlm modelspreferredBackend (Metal delegate now activated)addAudio + enableAudioModality)dart:io imports with conditional imports for WASM compilation supportexample/integration_test/benchmark_comparison_test.dart for comparing model performance on deviceToolChoice enum: auto / required / none parameter in createChat() to control tool calling behavior
auto / required / none parameter in createChat() to control tool calling behaviorParallelFunctionCallResponse for multiple function calls in one responseFunctionCallFormat implementations (Gemma, Qwen, DeepSeek, Llama, Phi, FunctionGemma)<tool_call> Format: Qwen/Mistral-style function call parsing<|tool_calls|> format supportnativeLibraryDir to LiteRT-LM Backend.NPU()Dual-Prefix Embeddings (TaskType): Improved RAG retrieval quality with query/document prefixes
TaskType.retrievalQuery (default) β for search queriesTaskType.retrievalDocument β for document indexingEmbedData.TaskType)addDocument() automatically uses document prefix.tflite embedding models (EmbeddingGemma, Gecko) on macOS, Windows, Linux
dart:ffi β no gRPC, no JVM overheaddart_sentencepiece_tokenizer (BPE + Unigram, auto-detect format)sqlite3 dart:ffi replacing platform-specific codepatrol dependency, migrated all integration tests to standard integration_testLiteRT-LM 0.9.0-beta: Updated from 0.9.0-alpha02 on Android and Desktop (JVM)
Conversation.cancelProcess()CancelGeneration RPCLlmInference.cancelProcessing() (MediaPipe 0.10.26)Conversation.cancelProcess()CancelGeneration RPCLlmInference.cancelProcessing() (MediaPipe 0.10.26)Pre-converted tokenizer.json files for iOS embedding models (avoids SentencePiece/TFLite protobuf conflict).
Pre-converted tokenizer.json files for iOS embedding models (avoids SentencePiece/TFLite protobuf conflict).
model.type field in tokenizer.jsoniosPath parameter for platform-aware tokenizer downloads
litertlm-server.jar URL to v0.12.5 (#189)package attribute from AndroidManifest.xmlAndroid ProGuard Fix: Added ProGuard rules for LiteRT-LM classes
Android LiteRT-LM Engine: Added LiteRT-LM inference engine for Android
.litertlm β LiteRT-LM, .task/.bin β MediaPipe)supportAudio parameter in session configurationModel Deletion Fix: Fixed model deletion not removing metadata
Web Large Model Support: WebStorageMode for models >2GB via OPFS streaming
WebStorageMode for models >2GB via OPFS streaming (#162)Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.16...v0.12.0
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.16...v0.12.0
.litertlm model format only (MediaPipe .task/.bin not supported on desktop)Fix iOS embedding crash on repeated inference (v0.11.16) by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/156
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.15...v0.11.16
Add Open In Colab badges to FunctionGemma pipeline links by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/149
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.14...v0.11.15
Feature/builder validation tdd by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/138
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.13...v0.11.14
ModelType.functionGemma for Google's specialized function calling model
0.11.7 by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/129
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.9...v0.11.13
@0.11.13/web/*.js)0.11.7 by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/129
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.9...v0.11.12
tasks-vision-image-generator to 0.10.26.1 for Android 15+ compatibilityπ Mobile Build Fix: Fixed compilation errors on iOS/Android platforms
β οΈ BREAKING CHANGE: Explicit initialization now required
await FlutterGemma.initialize() in main() before using the pluginFull Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.7...v0.11.9
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/v0.11.7...v0.11.9
π€ CI/CD Automation: Added GitHub Actions workflows for automated testing and release builds
0.10.6 clean by @DenisovAV in https://github.com/DenisovAV/flutter_gemma/pull/110
Full Changelog: https://github.com/DenisovAV/flutter_gemma/compare/0.10.6...v0.11.7
π iOS Simulator Fix: Fixed "Filename cannot contain path separators" crash on iOS Simulator
canResume(), resume(), cancel() methods from DownloadService interfaceβ οΈ Deprecated: Marked legacy asset/file management methods as deprecated with migration hints
New: Fluent builder API with FlutterGemma.installModel().fromNetwork/fromAsset/fromBundled/fromFile()
FlutterGemma.installModel().fromNetwork/fromAsset/fromBundled/fromFile()modelManager.downloadModelWithProgress()) still works as facadeπ Web Multimodal Support: Added full multimodal image processing support for web platform π MediaPipe 0.10.25: Updated to MediaPipe GenAI v0.10.25 for
π Web Multimodal Support: Added full multimodal image processing support for web platform
π MediaPipe 0.10.25: Updated to MediaPipe GenAI v0.10.25 for web compatibility
π¦ LiterTLM Format Support: Added support for .litertlm model files optimized for web platform
Download the demo APK below to test Flutter Gemma with:
𧬠Generated with Claude Code
π‘οΈ Fixed: Updated ProGuard rules for Android release build compatibility
π Fixed: Export missing ModelType and other public API types to resolve import issues
π Embedding Models Support: Added full support for text embedding models
InferenceModelSpec and EmbeddingModelSpec for better model organizationreplace and keepπ§ Model Replace Policy: Added configurable model replacement system with keep/replace options and ensureModelReady() method
ensureModelReady() methodHuggingFaceDownloader wrapper to handle CDN server inconsistencies and resume failures.task files (MediaPipe-handled) and .bin/.tflite files (manual formatting)π Stop Generation: Added Android support for stopping text generation mid-process with session.cancelRequestGenerationAsync() (#89, #19, #34)
session.cancelRequestGenerationAsync() (#89, #19, #34)π Documentation: Updated README with comprehensive model information and usage examples
π₯ Background Downloads: Added background download support for model files
π New Models: Added support for 4 new compact models:
π§ Thinking Mode: Added thinking mode support for DeepSeek models with persistent thinking bubbles
β¨ Function Calling: Added support for function calling, allowing models to interact with external tools.
πΌοΈ MULTIMODAL SUPPORT: Added full support for text + image input with Gemma 3 Nano vision models
Message class with support for text, image, and multimodal content
Message.text() - for text-only messagesMessage.withImage() - for text + image messagesMessage.imageOnly() - for image-only messagesmessage.hasImage - to check if message contains imageπ GEMMA 3 NANO SUPPORT: Added full support for Gemma 3 Nano models
input_pos != nullptr errorsUpgraded Mediapipe to 0.10.24 for iOS and Android
- iOS LoRA support added - iOS topP support added
- Readme updated
- Readme updated
Add web platform support in pubspec.yaml
Upgraded Mediapipe to 0.10.22 for Android and Web
Added Chat functionality for instruction tuned model
Added opportunity to manage inference session
Fixed crash on generation for Android
IMPORTANT: Breaking changes in the API
- Added close method
One column per month.