Skip to main content
Models are the foundation of Definable. The Model base class provides a consistent interface for invoking any supported LLM provider, whether synchronously, asynchronously, or via streaming.

​
Supported Providers

ProviderClassDefault ModelDocs
OpenAIOpenAIChatgpt-4oOpenAI
DeepSeekDeepSeekChatdeepseek-chatDeepSeek
MoonshotMoonshotChatkimi-k2-turbo-previewMoonshot
xAIxAIgrok-betaxAI
CustomOpenAILike—OpenAI-Like

​
Basic Usage

from definable.models import OpenAIChat

model = OpenAIChat(id="gpt-4o")
response = model.invoke(messages=[{"role": "user", "content": "Hello!"}])
print(response.content)

​
Common Parameters

All model classes accept these parameters:
​
id
str
required
The model identifier (e.g., "gpt-4o", "deepseek-chat").
​
api_key
str
API key for authentication. Defaults to the provider’s environment variable.
​
base_url
str
Override the API base URL. Useful for proxies or self-hosted endpoints.
​
temperature
float
Sampling temperature (0.0 to 2.0). Lower values are more deterministic.
​
max_tokens
int
Maximum number of tokens to generate.
​
timeout
float
Request timeout in seconds.
​
max_retries
int
Maximum number of retries on transient failures.

​
Invocation Methods

Every model supports four ways to call it:
MethodSync/AsyncStreamingReturns
invoke()SyncNoModelResponse
ainvoke()AsyncNoModelResponse
invoke_stream()SyncYesIterator[ModelResponse]
ainvoke_stream()AsyncYesAsyncIterator[ModelResponse]
response = model.invoke(messages=[{"role": "user", "content": "Hello!"}])
print(response.content)

​
Retry Configuration

All models support automatic retries with configurable backoff: 
model = OpenAIChat(
    id="gpt-4o",
    retries=3,
    delay_between_retries=1,
    exponential_backoff=True,
)
​
retries
int
default:"0"
Number of retry attempts on failure.
​
delay_between_retries
int
default:"1"
Base delay in seconds between retries.
​
exponential_backoff
bool
default:"false"
Whether to use exponential backoff for retries.

​
Response Caching

Cache model responses locally for development and testing: 
model = OpenAIChat(
    id="gpt-4o",
    cache_response=True,
    cache_ttl=3600,
    cache_dir=".cache/models",
)

​
Model Resolution from Strings

You can resolve a model from a "provider:model-id" string: 
from definable.models.utils import get_model

model = get_model("openai:gpt-4o")
model = get_model("deepseek:deepseek-chat")

​
Next Steps

Streaming

Stream responses token-by-token.

Structured Output

Return Pydantic models from LLM calls.