AutoGPT

AutoGPT is a powerful, recursive autonomous agent framework. Because AutoGPT can run in loops and consume significant tokens, routing it through Cencori is the safest way to deploy it — granting you strict spend caps and real-time observability.

Quick Start

1. Deploy from the Marketplace

Navigate to Agents → Agent Marketplace in your project dashboard and click Deploy Agent on the AutoGPT card.

2. Generate an API Key

Go to the agent's Configuration tab and click Generate Key. Copy the key.

3. Install AutoGPT

Follow the official AutoGPT installation instructions.

4. Configure .env

In your AutoGPT directory, find the .env file (or copy .env.template to .env) and update the OpenAI settings to point to Cencori:

# Point base URL to Cencori Gateway
OPENAI_API_BASE=https://cencori.com/api/v1
 
# Use your Cencori agent key
OPENAI_API_KEY=cake_YOUR_KEY_HERE
 
# Set models (must match models in your Cencori dashboard)
SMART_LLM=gpt-4o
FAST_LLM=gpt-4o-mini

5. Run

Run AutoGPT as normal.

autogpt run

All recursive loops, web searches, and file operations will now log to your Cencori dashboard. If AutoGPT gets stuck in a loop, Cencori will automatically kill the connection once it hits your project's spend limit.