Prompts API - Genkit

Prompts encapsulate model configuration and templates.

Loading Prompts

from genkit import Genkit

ai = Genkit(
    prompt_dir="./prompts",  # Auto-load .prompt files
)

Executing Prompts

# Get a loaded prompt
prompt = ai.registry.get_prompt("myPrompt")

# Execute
response = await prompt.execute(
    input={"name": "Alice"},
)

ExecutablePrompt

from genkit import ExecutablePrompt

prompt: ExecutablePrompt = ai.registry.get_prompt("greeting")

response = await prompt.execute(
    input={"name": "Bob"},
    config={"temperature": 0.8},
)

Prompt Configuration

Prompts are typically defined in .prompt files:

---
model: gemini-2.0-flash
input:
  schema:
    name: string
output:
  format: json
  schema:
    greeting: string
---
Say hello to {{ name }}

PromptGenerateOptions

from genkit import PromptGenerateOptions

options = PromptGenerateOptions(
    model="gemini-2.0-flash",
    config={"temperature": 0.7},
    tools=["my_tool"],
)

⌘I