On-device LLM

Model Size Function Calling Thinking Vision Multilingual URL Gemma 4 E2B 2.4GB ✅ ✅ ✅ ✅ litert-community/gemma-4-E2B-it-litert-lm/resolve/main/gemma-4-E2B-it.litertlm Gemma 4 E4B 4.3GB ✅ ✅ ✅ ✅ litert-community/gemma-4-E4B-it-litert-lm/resolve/main/gemma-4-E4B-it.litertlm Gemma3n E2B ~3GB ✅ ❌ ✅ ✅ google/gemma-3n-E2B-it-litert-lm/resolve/main/gemma-3n-E2B-it.litertlm Gemma3n E4B ~6GB ✅ ❌ ✅ ✅ google/gemma-3n-E4B-it-litert-lm/resolve/main/gemma-3n-E4B-it.litertlm FastVLM 0.5B 0.5GB ❌ ❌ ✅ ✅ litert-community/FastVLM-0.5B/resolve/main/FastVLM-0.5B.litertlm Gemma 3 1B 0.5GB ✅ ❌ ❌ ✅ litert-community/Gemma3-1B-IT/resolve/main/Gemma3-1B-IT_multi-prefill-seq_q4_ekv4096.litertlm Gemma 3 270M 0.3GB ❌ ❌ ❌ ✅ litert-community/gemma-3-270m-it/resolve/main/gemma-3-270m-it.litertlm FunctionGemma 270M 284MB ✅ ❌ ❌ ✅ sasha-denisov/function-gemma-270M-it/resolve/main/function-gemma-270M-it.litertlm Qwen3 0.6B 586MB ✅ ✅ ❌ ✅ litert-community/Qwen3-0.6B/resolve/main/Qwen3-0.6B.litertlm Qwen 2.5 0.5B 0.5GB ✅ ❌ ❌ ✅ litert-community/Qwen2.5-0.5B-Instruct/resolve/main/Qwen2.5-0.5B-Instruct.litertlm Qwen 2.5 1.5B 1.6GB ✅ ❌ ❌ ✅ litert-community/Qwen2.5-1.5B-Instruct/resolve/main/Qwen2.5-1.5B-Instruct.litertlm Phi-4 Mini 3.9GB ✅ ❌ ❌ ✅ litert-community/Phi-4-mini-instruct/resolve/main/Phi-4-mini-instruct.litertlm DeepSeek R1 1.7GB ✅ ✅ ❌ ✅ litert-community/DeepSeek-R1-Distill-Qwen-1.5B/resolve/main/DeepSeek-R1-Distill-Qwen-1.5B.litertlm SmolLM 135M 135MB ❌ ❌ ❌ English litert-community/SmolLM-135M-Instruct/resolve/main/SmolLM-135M-Instruct.litertlm

Refs:

• https://v-chandra.github.io/on-device-llms/
• https://pub.dev/packages/flutter_gemma