replicate-cli

Run AI model predictions and manage deployments via the Replicate CLI.

40|7|Updated Oct 20, 2025
One-click install
npx skills add https://github.com/rawveg/skillsforge-marketplace --skill replicate-cli
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: replicate-cli
Source: https://github.com/rawveg/skillsforge-marketplace/tree/main/replicate-cli
Command: npx skills add https://github.com/rawveg/skillsforge-marketplace --skill replicate-cli

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill guides you through using the Replicate CLI to run predictions, manage models and deployments, and fine-tune models from the terminal.

Core Features & Use Cases

  • Running Predictions: Use replicate run to execute models with inputs.
  • Models & Deployments: List, show, create, and manage models and deployments.
  • Training & Scaffolding: Fine-tune models and scaffold projects from existing predictions.

Quick Start

Run a simple image-generation model: replicate run stability-ai/sdxl prompt="a futuristic city" --save

Frequently Asked Questions about replicate-cli

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run AI model predictions from the command line using Replicate?

Use the replicate run command to execute models directly from your terminal. Specify the model name and input parameters—for example, replicate run stability-ai/sdxl prompt="a futuristic city" --save executes an image generation model and saves the output.

Can I manage model deployments and fine-tuning through the CLI?

Yes, the Replicate CLI lets you create, list, and manage deployments and models. You can also fine-tune models and scaffold projects from existing predictions using dedicated CLI commands.

What authentication and setup do I need to use the Replicate CLI?

You need to authenticate with Replicate through the CLI, which handles credential management. The Skill covers documented authentication workflows so you can securely run predictions and manage deployments.

How do I handle file uploads and streaming outputs when running predictions?

The Replicate CLI supports file uploads as inputs and streaming outputs for long-running predictions. Commands accept input files, and responses can stream results directly to your terminal or be saved for batch processing.

Can I inspect model schemas and select hardware for predictions?

Yes, the CLI includes commands to inspect model schemas so you understand required inputs. You can also specify hardware selection through flags when running predictions to optimize for your workload.

Does the Replicate CLI work for batch processing across multiple predictions?

The Replicate CLI supports batch processing workflows, allowing you to run multiple predictions efficiently. Combined with file uploads and output handling, it scales to process large input sets.