Anthropic - Chat Completion
Leverage advanced AI models from Anthropic to perform complex tasks such as categorization, analysis, summarization, or decision support.
For more information on the API for chat completion, see Create a Message (opens in a new tab).
SDK Import:
from admyral.actions import anthropic_chat_completionArguments:
| Argument Name | Description | Required |
|---|---|---|
Model model | The model to use for the chat completion (e.g., claude-3-5-sonnet-20240620). | Yes |
Prompt prompt | The input prompt to use for generating the chat completion. | Yes |
Top P top_p | Value between 0 and 1 for nucleus sampling. Only tokens with the top P probability mass are considered. Recommended to tweak this or temperature, but not both. | - |
Temperature temperature | Sampling temperature, between 0 and 2. Higher values (e.g., 0.8) increase randomness, while lower values (e.g., 0.2) make output more deterministic. Recommended to adjust this or Top P, but not both. | - |
Max Tokens max_tokens | The maximum number of tokens to generate for the completion. | - |
Stop Tokens stop_tokens | A list of tokens to stop the completion upon encountering. | - |
Returns
A string.
Required Secrets
| Secret Placeholder | Description |
|---|---|
ANTHROPIC_SECRET | Anthropic secret. See Anthropic setup |
SDK Example
response = anthropic_chat_completion(
model="claude-3-5-sonnet-20240620",
prompt="Why is the ocean salty?",
secrets={
"ANTHROPIC_SECRET": "my_stored_anthropic_secret"
}
)Example Output:
{
"content": [
{
"text": "Hi! My name is Claude.",
"type": "text"
}
],
"id": "msg_013Zva2CMHLNnXjNJJKqJ2EF",
"model": "claude-3-5-sonnet-20240620",
"role": "assistant",
"stop_reason": "end_turn",
"stop_sequence": null,
"type": "message",
"usage": {
"input_tokens": 2095,
"output_tokens": 503
}
}