Skip to main content

CODITECT Gemini API Executor

Multi-model AI execution framework - Gemini (Google) provider implementation. Provides standardized interface for executing prompts against Gemini models with streaming support, token tracking, and error handling.

Features:

  • Gemini 1.5 Pro, Gemini 1.5 Flash support
  • Streaming and batch response modes
  • Token usage tracking and cost estimation
  • Automatic retry with exponential backoff
  • Safety settings configuration
  • Multimodal input support (text + images)
  • Context window management (1M+ tokens)

Usage: from core.execute_gemini import GeminiExecutor

executor = GeminiExecutor(model="gemini-1.5-pro")
response = executor.execute("Explain quantum computing")

# Streaming
for chunk in executor.stream("Write a poem"):
    print(chunk, end="")

Environment Variables: GOOGLE_API_KEY - Required API key GEMINI_MODEL - Default model (optional)

Author: AZ1.AI INC. Framework: CODITECT Copyright: © 2025 AZ1.AI INC. All rights reserved.

File: execute_gemini.py

Classes​

ExecutionResult​

Result from Gemini execution.

GeminiExecutor​

Gemini API executor with standardized interface.

Functions​

main()​

CLI interface for Gemini executor.

to_dict()​

No description

execute(prompt, system, history, max_tokens, temperature)​

Execute prompt against Gemini model.

stream(prompt, system, max_tokens)​

Stream response from Gemini model.

Usage​

python execute_gemini.py