Replicate AI

Replicate AI

The robomotion replicateai CLI runs AI model predictions on Replicate's cloud infrastructure. It supports running models synchronously or asynchronously, streaming predictions, listing and searching models, and managing prediction lifecycle.

When to use

• Run AI model predictions (image gen, LLMs, audio, video)
• Submit async predictions and retrieve results later
• Stream prediction outputs in real-time
• Search and discover Replicate models

Prerequisites

• robomotion CLI installed
• Package installed: robomotion install replicateai
• Replicate API token configured via Robomotion vault

IMPORTANT: Session Mode Required for Multi-Step Operations

Each CLI command runs as a separate process — connection IDs from connect do NOT persist across calls. You MUST use --session on connect and pass --session-id to all subsequent commands.

Workflow

1. Install (once): robomotion install replicateai
2. Connect with session:

   robomotion replicateai replicateai_connect --session --output json
   # → {"outClientId":"<client-id>","session_id":"<session-id>"}
   

3. Use the returned client-id and session-id in all subsequent commands:

   robomotion replicateai replicateai_run --client-id "<client-id>" --model <owner/model> --input <json> --session-id "<session-id>" --output json
   

4. Disconnect when done:

   robomotion replicateai replicateai_disconnect --client-id "<client-id>" --session-id "<session-id>" --output json
   

Always append --output json to get structured JSON results.

Commands Reference

• robomotion replicateai connect --session --output json Connect to Replicate AI and create a client session for running machine learning models
• robomotion replicateai disconnect --connection-id --session-id "<session-id>" --output json Closes the Replicate AI connection and releases resources
• robomotion replicateai run_model --connection-id --owner/model:version --input [--timeout-(seconds)] --session-id "<session-id>" --output json Run a machine learning model on Replicate AI and wait for completion
• robomotion replicateai get_prediction --connection-id --prediction-id [--timeout-(seconds)] --session-id "<session-id>" --output json Get the status and output of a prediction by ID
• robomotion replicateai create_prediction --connection-id --owner/model:version --input [--timeout-(seconds)] --session-id "<session-id>" --output json Create a prediction without waiting for completion - use Get Prediction to check status
• robomotion replicateai get_model --connection-id --model-owner --model-name [--timeout-(seconds)] --session-id "<session-id>" --output json Get information about a model including description٫ run count٫ and URLs
• robomotion replicateai create_training --connection-id --model-owner --model-name --version --destination --inputs --session-id "<session-id>" --output json Create a training job to fine-tune a model on your own data
• robomotion replicateai get_training --connection-id --training-id --session-id "<session-id>" --output json Get the status and output of a training job by ID
• robomotion replicateai list_predictions --connection-id --session-id "<session-id>" --output json List recent predictions from your Replicate account
• robomotion replicateai cancel_prediction --connection-id --prediction-id --session-id "<session-id>" --output json Cancel a running prediction by ID

Environment

• ROBOMOTION_API_TOKEN (if vault credentials are needed)