Gemini.Types.Response.Model (GeminiEx v0.2.1)

Model information response structure.

Represents the complete model metadata returned by the Gemini API, including capabilities, token limits, and generation parameters.

Summary

Types

capability_level()

Model capability level based on features and limits

capacity_tier()

Token capacity classification

t()

Functions

capabilities_summary(model)

Generate a comprehensive capabilities summary.

capability_score(model)

Calculate a capability score for model comparison.

compare_capabilities(model1, model2)

Compare two models by capability.

effective_base_id(model)

Get the effective base model ID.

has_advanced_params?(model)

Check if model has advanced generation parameters.

input_capacity_tier(model)

Classify model's input token capacity.

latest_version?(model)

Check if this appears to be the latest version of a model.

model_family(model)

Extract model family from the base model ID.

output_capacity_tier(model)

Classify model's output token capacity.

production_ready?(model)

Determine if model is suitable for production use.

supports_embeddings?(model)

Check if model supports embeddings.

supports_method?(model, method)

Check if model supports a specific generation method.

supports_streaming?(model)

Check if model supports streaming content generation.

supports_token_counting?(model)

Check if model supports token counting.

Types

capability_level()

@type capability_level() :: :basic | :standard | :advanced | :premium

Model capability level based on features and limits

capacity_tier()

@type capacity_tier() :: :small | :medium | :large | :very_large

Token capacity classification

t()

@type t() :: %Gemini.Types.Response.Model{
  base_model_id: String.t(),
  description: String.t(),
  display_name: String.t(),
  input_token_limit: integer(),
  max_temperature: float() | nil,
  name: String.t(),
  output_token_limit: integer(),
  supported_generation_methods: [String.t()],
  temperature: float() | nil,
  top_k: integer() | nil,
  top_p: float() | nil,
  version: String.t()
}

Functions

capabilities_summary(model)

@spec capabilities_summary(t()) :: map()

Generate a comprehensive capabilities summary.

Example Response

%{
  supports_streaming: true,
  supports_token_counting: true,
  supports_embeddings: false,
  has_temperature: true,
  has_top_k: true,
  has_top_p: false,
  method_count: 3,
  input_capacity: :very_large,
  output_capacity: :medium
}

capability_score(model)

@spec capability_score(t()) :: integer()

Calculate a capability score for model comparison.

Higher scores indicate more capable models.

compare_capabilities(model1, model2)

@spec compare_capabilities(t(), t()) :: :lt | :eq | :gt

Compare two models by capability.

Returns :lt, :eq, or :gt based on capability scores.

effective_base_id(model)

@spec effective_base_id(t()) :: String.t()

Get the effective base model ID.

Prefers the base_model_id field, but falls back to extracting from the name if base_model_id is nil.

Examples

iex> Model.effective_base_id(%Model{base_model_id: "gemini-2.0-flash-lite"})
"gemini-2.0-flash-lite"

iex> Model.effective_base_id(%Model{name: "models/gemini-2.5-pro", base_model_id: nil})
"gemini-2.5-pro"

has_advanced_params?(model)

@spec has_advanced_params?(t()) :: boolean()

Check if model has advanced generation parameters.

Returns true if the model supports temperature, top_p, or top_k parameters.

input_capacity_tier(model)

@spec input_capacity_tier(t()) :: capacity_tier()

Classify model's input token capacity.

Examples

iex> Model.input_capacity_tier(%Model{input_token_limit: 2_000_000})
:very_large

iex> Model.input_capacity_tier(%Model{input_token_limit: 30_000})
:medium

latest_version?(model)

@spec latest_version?(t()) :: boolean()

Check if this appears to be the latest version of a model.

Heuristic based on name patterns (no version suffix, "latest" in name).

model_family(model)

@spec model_family(t()) :: String.t()

Extract model family from the base model ID.

Examples

iex> Model.model_family(%Model{base_model_id: "gemini-2.0-flash-lite"})
"gemini"

iex> Model.model_family(%Model{base_model_id: "text-embedding-004"})
"text"

output_capacity_tier(model)

@spec output_capacity_tier(t()) :: capacity_tier()

Classify model's output token capacity.

production_ready?(model)

@spec production_ready?(t()) :: boolean()

Determine if model is suitable for production use.

Based on capability, capacity, and stability indicators.

supports_embeddings?(model)

@spec supports_embeddings?(t()) :: boolean()

Check if model supports embeddings.

supports_method?(model, method)

@spec supports_method?(t(), String.t()) :: boolean()

Check if model supports a specific generation method.

Examples

iex> Model.supports_method?(model, "generateContent")
true

iex> Model.supports_method?(model, "nonexistentMethod")
false

supports_streaming?(model)

@spec supports_streaming?(t()) :: boolean()

Check if model supports streaming content generation.

supports_token_counting?(model)

@spec supports_token_counting?(t()) :: boolean()

Check if model supports token counting.