🔥 DeepEval 4.0 just got released. Read the announcement.
Evaluation Models

Moonshot

DeepEval's integration with Moonshot AI allows you to use any Moonshot models to power all of DeepEval's metrics.

Command Line

To configure your Moonshot model through the CLI, run the following command:

deepeval set-moonshot \
    --model="kimi-k2-0711-preview" \
    --temperature=0

Python

Alternatively, you can define KimiModel directly in python code:

from deepeval.models import KimiModel
from deepeval.metrics import AnswerRelevancyMetric

model = KimiModel(
    model="kimi-k2-0711-preview",
    api_key="your-api-key",
    temperature=0
)

answer_relevancy = AnswerRelevancyMetric(model=model)

To use any Moonshot model directly in deepeval, set the USE_MOONSHOT_MODEL=1 in your env and simply pass the name of your desired model in your metric initialization:

from deepeval.metrics import AnswerRelevancyMetric

answer_relevancy = AnswerRelevancyMetric(
    model="kimi-k2-0711-preview",
)

You should also set the other necessary vars like MOONSHOT_API_KEY to be able to use the Moonshot models as shown above.

There are ZERO mandatory and SIX optional parameters when creating an KimiModel:

  • [Optional] model: A string specifying the name of the Kimi model to use. Defaults to MOONSHOT_MODEL_NAME if not passed; raises an error at runtime if unset.
  • [Optional] api_key: A string specifying your Kimi API key for authentication. Defaults to MOONSHOT_API_KEY if not passed; raises an error at runtime if unset.
  • [Optional] temperature: A float specifying the model temperature. Defaults to TEMPERATURE if not passed; falls back to 0.0 if unset and raises if < 0.
  • [Optional] cost_per_input_token: A float specifying the cost for each input token for the provided model. Defaults to MOONSHOT_COST_PER_INPUT_TOKEN if available in deepeval's model cost registry, else None.
  • [Optional] cost_per_output_token: A float specifying the cost for each output token for the provided model. Defaults to MOONSHOT_COST_PER_OUTPUT_TOKEN if available in deepeval's model cost registry, else None.
  • [Optional] generation_kwargs: A dictionary of additional generation parameters forwarded to the OpenAI chat.completions.create(...) call.

Available Moonshot Models

Below is a comprehensive list of available Moonshot models:

  • kimi-k2-0711-preview
  • kimi-thinking-preview
  • moonshot-v1-8k
  • moonshot-v1-32k
  • moonshot-v1-128k
  • moonshot-v1-8k-vision-preview
  • moonshot-v1-32k-vision-preview
  • moonshot-v1-128k-vision-preview
  • kimi-latest-8k
  • kimi-latest-32k
  • kimi-latest-128k

On this page