Skip to main content

Model

What it is​

  • Lightweight data holders for LangChain chat models and related metadata.
  • Provides a base Model class and a ChatModel specialization with context_window and a model_type.

Public API​

Enums​

  • ModelType
    • CHAT: Identifies chat-capable models.

Classes​

  • Model

    • Purpose: Store a langchain_core BaseChatModel instance plus identifying and descriptive metadata.
    • Constructor:
      • Model(model_id: str, provider: str, model: BaseChatModel, ..., default_parameters: Optional[dict] = None)
    • Key attributes (all set directly on the instance):
      • Required: model_id, provider, model
      • Optional metadata: name, owner, description, image, created_at, canonical_slug, hugging_face_id
      • Optional structured info: pricing, architecture, top_provider, per_request_limits, supported_parameters, default_parameters

  • ChatModel(Model)

    • Purpose: A Model with chat-specific fields.
    • Class attributes:
      • model_type: ModelType = ModelType.CHAT
    • Additional instance attributes:
      • context_window: Optional[int] - maximum context length in tokens
    • Constructor:
      • ChatModel(model_id: str, provider: str, model: BaseChatModel, context_window: Optional[int] = None, ..., default_parameters: Optional[dict] = None)

Configuration/Dependencies​

  • Depends on:
    • langchain_core.language_models.chat_models.BaseChatModel (the wrapped chat model type)
    • pydantic.Field and typing.Annotated (used for type annotations/metadata only)
    • Standard library: datetime, enum, typing.Optional

Usage​

from datetime import datetime
from naas_abi_core.models.Model import ChatModel

# `model` must be an actual LangChain BaseChatModel instance in real usage.
# Here we use a placeholder to show construction only.
model = object()  # replace with a real BaseChatModel

chat_model = ChatModel(
    model_id="gpt-4.1",
    provider="openai",
    model=model,  # should be a BaseChatModel
    context_window=128000,
    name="GPT-4.1",
    created_at=datetime.utcnow(),
)

print(chat_model.model_id)
print(chat_model.model_type)
print(chat_model.context_window)

Caveats​

  • These classes are not Pydantic BaseModels; the Field(...) metadata is not automatically enforced/validated at runtime.
  • model is annotated as BaseChatModel, but there is no runtime type checking in __init__.