Skip to main content

​
Installation

pip install reminix

​
Usage

from reminix import Reminix

client = Reminix(api_key="your-api-key")

​
Async Client

The SDK ships with a fully async client for use in async applications. 
from reminix import AsyncReminix

client = AsyncReminix(api_key="your-api-key")

response = await client.agents.invoke("my-agent", input={
    "prompt": "Hello",
})

​
Configuration

​
api_key
str
required
Your project API key or personal access token. Can also be set via the REMINIX_API_KEY environment variable.
​
base_url
str
API base URL. Defaults to https://api.reminix.com/v1.
​
timeout
float
Request timeout in seconds. Defaults to 60.
​
default_headers
dict
Headers included in every request. Use for PAT authentication with the X-Project header.

​
Environment Variables

You can set your API key as an environment variable instead of passing it directly. 
export REMINIX_API_KEY="reminix_sk_..."

# API key auto-detected from environment
client = Reminix()
When the api_key parameter is omitted, the SDK automatically reads from the REMINIX_API_KEY environment variable.

​
Framework Packages

Reminix provides optional integration packages for popular AI frameworks. Install them alongside the base SDK to use framework-specific helpers with your Reminix agents.
PackageFrameworkAlso in TypeScript?
reminix-openaiOpenAI integrationYes
reminix-anthropicAnthropic integrationYes
reminix-langchainLangChain integrationYes
reminix-googleGoogle AI integrationYes
The Vercel AI SDK integration (@reminix/vercel-ai) is TypeScript-only.
Install any combination you need: 
pip install reminix reminix-openai reminix-anthropic
Framework packages are optional. The base reminix SDK is all you need to invoke and chat with agents.

​
Next steps

invoke()

Call task agents and stream results.

chat()

Build multi-turn conversations.

Creating Agents

Define and serve agents from your own code.

Error Handling

Catch typed exceptions and add retries.