Files
elasticsearch-js/docs/reference/api/InferenceAnthropicServiceSettings.md
2025-05-28 10:02:30 -05:00

602 B

InferenceAnthropicServiceSettings [interface-InferenceAnthropicServiceSettings]

Name Type Description
api_key string A valid API key for the Anthropic API.
model_id string The name of the model to use for the inference task. Refer to the Anthropic documentation for the list of supported models.
rate_limit InferenceRateLimitSetting This setting helps to minimize the number of rate limit errors returned from Anthropic. By default, the anthropic service sets the number of requests allowed per minute to 50.