venice_ai.utils.models

Model filtering and ID building helpers.

Exports: * :func:get_filtered_models — filter a list of models by capabilities * :func:_apply_model_filters — internal filter implementation * :func:build_model_id — construct a Venice model ID with feature suffixes

get_filtered_models

def get_filtered_models(models: Sequence[ModelDetails],
                        model_type: str | None = None,
                        supports_vision: bool | None = None,
                        supports_reasoning: bool | None = None,
                        supports_function_calling: bool | None = None,
                        supports_web_search: bool | None = None,
                        supports_log_probs: bool | None = None,
                        optimized_for_code: bool | None = None,
                        quantization: str | None = None,
                        is_beta: bool | None = None,
                        has_trait: str | None = None) -> list[ModelDetails]

Filters a list of models based on various capabilities.

Arguments:

• models: A list of model details objects to filter.
• model_type: Optional. Filter for model type.
• supports_vision: Optional. Filter by vision support.
• supports_reasoning: Optional. Filter by reasoning support.
• supports_function_calling: Optional. Filter by function calling support.
• supports_web_search: Optional. Filter by web search support.
• supports_log_probs: Optional. Filter by log probability support.
• optimized_for_code: Optional. Filter by code optimization.
• quantization: Optional. Filter by quantization type (e.g., "fp16", "fp8").
• is_beta: Optional. Filter by beta status.
• has_trait: Optional. Filter by a specific model trait.

Returns:

A new list of model details objects that match the filters.

build_model_id

def build_model_id(model: str, **params: str | int | float | bool) -> str

Build a Venice model ID with feature suffixes.

The Venice API supports model feature suffixes in the format: model_id:param=value&param2=value2

Arguments:

• model: Base model ID (e.g., "llama-3.3-70b").
• **params: Feature parameters (e.g., reasoning_effort="high").

Raises:

• ValueError: If model is empty or contains a ':' character (i.e., it already has a suffix).

Returns:

Model ID with feature suffix, or just the base model ID if no params are provided.