ai.analysis.model package¶

Submodules¶

ai.analysis.model.assistant_chat_model module¶

class ai.analysis.model.assistant_chat_model.AssistantChatModel(*values)[source]¶

Bases: Enum

Enum of supported assistant chat model identifiers.

GPT_4O¶

Identifier for the full GPT-4O model.

Type:: str

GPT_4_1¶

Identifier for the GPT-4.1 model.

Type:: str

GPT_4_1_MINI¶

Identifier for the GPT-4.1 Mini model.

Type:: str

GPT_4_1_NANO¶

Identifier for the GPT-4.1 Nano model.

Type:: str

GPT_4O_MINI¶

Identifier for the GPT-4O Mini model.

Type:: str

QWEN_2_5_7B¶

Identifier for the Qwen 2.5-7B Instruct model.

Type:: str

QWEN_2_5_7B_TOGETHER¶

Identifier for the Qwen 2.5-7B Instruct Turbo model.

Type:: str

LLAMA_70B¶

Identifier for the Meta LLaMA-3.3-70B Instruct model.

Type:: str

LLAMA_3B¶

Identifier for the Meta LLaMA-3.2-3B Instruct model.

Type:: str

GPT_4O = 'gpt-4o'¶

GPT_4_1 = 'gpt-4.1'¶

GPT_4_1_MINI = 'gpt-4.1-mini'¶

GPT_4_1_NANO = 'gpt-4.1-nano'¶

GPT_4O_MINI = 'gpt-4o-mini'¶

QWEN_2_5_7B = 'qwen/qwen-2.5-7b-instruct'¶

QWEN_2_5_7B_TOGETHER = 'Qwen/Qwen2.5-7B-Instruct-Turbo'¶

DEEP_SEEK_R1 = 'deepseek/deepseek-r1'¶

LLAMA_70B = 'meta-llama/llama-3.3-70b-instruct'¶

LLAMA_3B = 'meta-llama/llama-3.2-3b-instruct'¶

ai.analysis.model.token_price_calculator module¶

class ai.analysis.model.token_price_calculator.ModelPriceInfo(input_token_cost: Money, output_token_cost: Money)[source]¶

Bases: object

Cost information for input and output tokens of a model.

input_token_cost¶

Cost per input token.

Type:: Money

output_token_cost¶

Cost per output token.

Type:: Money

input_token_cost: Money¶

output_token_cost: Money¶

get_cost_for(token_type: TokenType) → Money[source]¶

Return the cost for the specified token type.

Parameters:: token_type (TokenType) – The type of token (INPUT or OUTPUT).
Returns:: The cost associated with the token type.
Return type:: Money
Raises:: ValueError – If an unsupported token type is provided.

__init__(input_token_cost: Money, output_token_cost: Money) → None¶

class ai.analysis.model.token_price_calculator.TokenPriceCalculator[source]¶

Bases: object

Calculator for determining token costs of supported assistant models.

model_price_infos: dict[AssistantChatModel, ModelPriceInfo] = {AssistantChatModel.DEEP_SEEK_R1: ModelPriceInfo(input_token_cost=$5e-07, output_token_cost=$2.15e-06), AssistantChatModel.GPT_4O: ModelPriceInfo(input_token_cost=$2.5e-06, output_token_cost=$1e-05), AssistantChatModel.GPT_4O_MINI: ModelPriceInfo(input_token_cost=$1.5e-07, output_token_cost=$6e-07), AssistantChatModel.GPT_4_1: ModelPriceInfo(input_token_cost=$2e-06, output_token_cost=$8e-06), AssistantChatModel.GPT_4_1_MINI: ModelPriceInfo(input_token_cost=$4e-07, output_token_cost=$1.6e-06), AssistantChatModel.GPT_4_1_NANO: ModelPriceInfo(input_token_cost=$1e-07, output_token_cost=$4e-07), AssistantChatModel.LLAMA_3B: ModelPriceInfo(input_token_cost=$1.5e-08, output_token_cost=$2.5e-08), AssistantChatModel.LLAMA_70B: ModelPriceInfo(input_token_cost=$1.2e-07, output_token_cost=$3e-07), AssistantChatModel.QWEN_2_5_7B: ModelPriceInfo(input_token_cost=$2.5e-08, output_token_cost=$5e-08), AssistantChatModel.QWEN_2_5_7B_TOGETHER: ModelPriceInfo(input_token_cost=$3e-07, output_token_cost=$3e-07)}¶

get_cost(model: AssistantChatModel, token_type: TokenType) → Money[source]¶

Calculate the cost for a given model and token type.

Parameters:

model (AssistantChatModel) – The chat model to query.
token_type (TokenType) – The type of token (INPUT or OUTPUT).

Returns:

The cost for one token of the specified type for the given model.

Return type:

Money

Raises:

ValueError – If the model is not found in the price information.

ai.analysis.model.token_type module¶

class ai.analysis.model.token_type.TokenType(*values)[source]¶