Package-level declarations
Types
Provider options for AlibabaProvider.embeddingModel — pass under the
Immutable byte container for LiteRT media parts.
Reasoning or side-channel delimiter configuration for LiteRT prompt templates.
Builder for LiteRTChannel.
Content value exchanged with a LiteRTConversation.
Host bridge for one LiteRT conversation/session.
Factory that creates a host LiteRT conversation for one SDK model call.
Prepared call delivered to LiteRTConversationFactory.create.
Builder for LiteRTConversationRequest.
LanguageModel adapter for on-device LiteRT-LM engines.
Settings that control LiteRT prompt/request mapping.
Builder for LiteRTLanguageModelSettings.
Message exchanged with a LiteRTConversation.
Builder for LiteRTMessage.
LiteRT-side message role used by the host conversation bridge.
Provider-options payload consumed by LiteRTLanguageModel.
LiteRT sampling configuration passed to the host engine.
Builder for LiteRTSamplerConfig.
How LiteRT streaming text snapshots should be interpreted.
Tool call emitted by a LiteRT conversation.
Builder for LiteRTToolCall.
Test stub for a LanguageModel. Drives the agent loop deterministically by replaying a list of ScriptedResponses — one per stream call.
One scripted response — what a single stream call should emit.
Properties
Functions
PascalCase factory — mirrors OpenAI(client, settings).
PascalCase factory — mirrors the OpenAI(...) reference faux-constructor.
PascalCase factory — mirrors the reference OpenAI(...) faux-constructor.
PascalCase factory — mirrors the OpenAI reference pattern.
PascalCase factory — mirrors OpenAI(...).
PascalCase factory — mirrors the OpenAI(...) reference faux-constructor.
PascalCase factory — mirrors the OpenAI(...) reference faux-constructor.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors OpenAI(...).
PascalCase factory — mirrors OpenAI(...), the Layer-3 reference pattern.
PascalCase factory — mirrors the OpenAI reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference faux-constructor.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors OpenAI(...).
PascalCase factory — mirrors the OpenAI reference pattern.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
Builds a LiteRT channel delimiter configuration.
Builds a prepared LiteRT conversation request.
Builds LiteRT language model settings.
Builds a LiteRT message.
Builds a LiteRT sampler configuration.
Builds a LiteRT tool call value.
PascalCase factory — mirrors the OpenAI(...) reference pattern.
PascalCase factory — mirrors OpenAI(...).
PascalCase factory mirroring OpenAI.
A model that just emits the given text once and finishes.
A model whose first call requests a tool, second call (after the tool result is appended) returns a final text response.
Convenience for a tool-call input expressed as JSON literal map.
PascalCase factory — the reference pattern that Layer-8 providers will follow.
Builds a provider for OpenAI-shaped HTTP APIs.
PascalCase factory — mirrors the reference OpenAI(...) faux-constructor.
PascalCase factory — mirrors OpenAI(...).
PascalCase factory — mirrors the OpenAI provider pattern.
PascalCase factory mirroring the OpenAI reference pattern.
PascalCase factory — mirrors the OpenAI reference pattern.
PascalCase factory — mirrors OpenAI(...).