Flutter Edge AI with GGUF
Where the upstream flutter_edge_ai_llama connector stands and which llamadart version it requires.
On this page
Flutter Edge AI owns installation, active-model selection and app chat lifecycle.
Its optional flutter_edge_ai_llama
provider uses llamadart for GGUF generation. The connector lives in
DenisovAV/flutter_edge_ai
and is not part of this repository.
Version boundary#
flutter_edge_ai_llama 0.0.4 is on pub.dev and requires exactly
llamadart >=0.15.1 <0.15.2, so it does not resolve alongside llamadart 0.17.
For the connector, follow its
README
and platform guide,
including its llamadart-specific setup; llamadart's build hook settings belong in
the application root.
To use llamadart 0.17 features, call llamadart directly or through Genkit until the connector widens its constraint.
Combining with LiteRT-LM#
Select llamadart_native_runtimes: [llama_cpp] when flutter_edge_ai_litertlm
is also in the app. Loading llamadart's LiteRT-LM distribution alongside it can
cause native symbol and library collisions. See
Alongside Flutter Edge AI.