vision-agents-plugins-openrouter 0.3.3

OpenRouter plugin for vision agents

OpenRouter Plugin

This plugin provides LLM capabilities using OpenRouter's API, which offers access to multiple LLM providers through a unified OpenAI-compatible interface. It enables developers to easily switch between different models from various providers (Anthropic, Google, OpenAI, etc.) without changing their code.

Features

• Access to multiple LLM providers through a single API
• OpenAI-compatible interface for easy integration
• Support for various models including Claude, Gemini, GPT, and more
• Automatic conversion of instructions to system messages
• Manual conversation history management

Installation

uv add vision-agents[openrouter]

Usage

from vision_agents.core import User, Agent
from vision_agents.plugins import openrouter, getstream, elevenlabs, deepgram, smart_turn

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="OpenRouter AI"),
    instructions="Be helpful and friendly to the user",
    llm=openrouter.LLM(
        model="anthropic/claude-haiku-4.5",
    ),
    tts=elevenlabs.TTS(),
    stt=deepgram.STT(),
    turn_detection=smart_turn.TurnDetection(),
)

Configuration

Parameter	Description	Accepted Values
api_key	OpenRouter API key	str | None. If not provided, uses OPENROUTER_API_KEY environment variable
base_url	OpenRouter API base URL	str. Default: "https://openrouter.ai/api/v1"
model	Model identifier to use	str. Default: "openrouter/andromeda-alpha". Examples: "anthropic/claude-haiku-4.5", "google/gemini-2.5-flash", "openai/gpt-4o"
**kwargs	Additional arguments passed to OpenAI LLM	Any additional parameters supported by the underlying OpenAI LLM implementation

Dependencies

• vision-agents
• vision-agents-plugins-openai