OpenClaw TUI Interaction

1. Course Content

Course Overview

OpenClaw TUI is a terminal-based text interaction interface. No graphical desktop environment is required - you can chat with the OpenClaw robot directly through the command line. TUI interaction is especially suitable for scenarios where you connect to the robot remotely via SSH and perform text-based interactions in the terminal.

Learning Objectives

• Master the method to start a terminal interaction session using the openclaw tui command

• Learn advanced usage of starting TUI with parameters

• Become familiar with common conversation commands (slash commands) in the TUI terminal

2. Starting TUI Conversation

Applicable Scenarios

TUI conversation is mainly applicable to the following scenarios:

• SSH Remote Connection — After logging into the robot remotely via SSH, you can directly chat with OpenClaw in the terminal

• Lightweight Interaction — Compared to Web UI, TUI consumes fewer resources and starts faster, making it suitable for quick debugging and daily use

Enter the following command directly in the terminal to start the TUI conversation interface:

openclaw tui

3. Common TUI Terminal Conversation Commands

• Conversation commands in the TUI terminal are exactly the same as slash commands in WebChat, with identical usage. All commands must start with / and be sent as independent messages

4. Common Errors and Solutions

4.1 Invalid API

Symptoms: After sending a message, the robot reports invalid API-KEY, authentication failure, or returns a 401 error.

Possible causes and solutions:

• API-KEY configuration error — Check if the API-KEY configuration item in openclaw.json is correct; reconfigure and restart the gateway

• API-KEY expired — Go to the model service provider's platform to check if the API-KEY has expired; generate a new KEY and update the configuration

• Model service provider address error — Confirm that the API address (base URL) for the model in openclaw models list is filled in correctly

xxxxxxxxxx
# View current model configuration
openclaw models list
# View model status
openclaw models status

4.2 Model Quota Exceeded

Symptoms: After sending a message, it reports that the model's call quota limit has been reached (Rate Limit / Quota Exceeded), or returns a 429 status code.

Possible causes and solutions:

• Quota exhausted — Model service providers have daily/hourly call limits or token limits for free/trial accounts; please check your account quota status

• Multiple concurrent sessions consuming quota — Having multiple sessions open for parallel conversations will accelerate quota consumption; it is recommended to reduce the number of concurrent sessions

• Switch model — If the current model's quota is limited, you can use the /model command to switch to another configured model to continue the conversation