Text rendering artifacts after long sessions: Run Cmd+K to clear the screen, or restart the Ghostty tab. This is rare but can happen with 100K+ line scrollback. Claude Code output truncated: Increase scrollback-limit in your config to 500000 if you regularly generate very long outputs. Shell ...

How to Set Up Claude Code in Ghostty (2026)

Last updated: April 21, 2026

The Workflow

Set up the Ghostty terminal emulator as a high-performance environment for Claude Code sessions. Ghostty’s GPU-accelerated rendering handles long AI-generated outputs without lag, making it ideal for extended coding sessions.

Expected time: 15 minutes Prerequisites: Ghostty 1.0+, Claude Code CLI installed, macOS or Linux

Setup

1. Install Ghostty

# macOS via Homebrew
brew install --cask ghostty
# Linux (build from source)
git clone https://github.com/ghostty-org/ghostty.git
cd ghostty
zig build -Doptimize=ReleaseFast
sudo cp zig-out/bin/ghostty /usr/local/bin/

Ghostty uses the GPU for text rendering, which prevents terminal freezes during large Claude Code outputs.

2. Create Ghostty Configuration

mkdir -p ~/.config/ghostty
cat > ~/.config/ghostty/config << 'EOF'
# Font configuration for code readability
font-family = JetBrains Mono
font-size = 14
font-thicken = true
# Window settings
window-padding-x = 12
window-padding-y = 8
window-decoration = true
window-theme = dark
# Performance settings for long Claude Code outputs
scrollback-limit = 100000
clipboard-read = allow
clipboard-write = allow
clipboard-paste-protection = false
# Colors (Catppuccin Mocha for reduced eye strain)
background = 1e1e2e
foreground = cdd6f4
cursor-color = f5e0dc
selection-background = 45475a
selection-foreground = cdd6f4
# Shell integration
shell-integration = zsh
shell-integration-features = cursor,sudo,title
# Keybindings for Claude Code workflow
keybind = super+shift+c=new_tab
keybind = super+k=clear_screen
keybind = super+shift+v=paste_from_clipboard
EOF

This config provides a comfortable environment optimized for reading AI-generated code.

3. Configure Shell Environment for Claude Code

cat >> ~/.zshrc << 'EOF'
# Claude Code in Ghostty
export ANTHROPIC_API_KEY="sk-ant-your-key-here"
export CLAUDE_CODE_THEME="dark"
# Aliases for quick access
alias cc="claude"
alias ccp="claude --print"
alias ccr="claude --resume"
# Function to start Claude Code in a named session
ccn() {
    echo "Starting Claude Code session: $1"
    claude --resume "$1" 2>/dev/null || claude
}
EOF
source ~/.zshrc

The aliases reduce typing for frequent Claude Code operations.

4. Set Up Split Panes for Code + Claude

# Ghostty supports native splits
# Use these keybindings (default):
# Ctrl+Shift+Enter - Split horizontally
# Ctrl+Shift+| - Split vertically
# Ctrl+Shift+Arrow - Navigate between panes

Run your editor in one pane and Claude Code in the adjacent pane for side-by-side development.

5. Verify

ghostty --version
# Expected output:
# ghostty 1.x.x
claude --version
# Expected output:
# claude-code x.x.x
claude --print "Say hello from Ghostty"
# Expected output:
# Hello from Ghostty! (or similar response)

Usage Example

A typical workflow with Claude Code in Ghostty using split panes:

# Terminal split: Left pane = editor, Right pane = Claude Code
# In right pane, start Claude Code
claude
# Generate a utility function
> Write a TypeScript debounce utility with proper generics and
> cancellation support. Include JSDoc comments.

Claude Code outputs directly in the Ghostty pane:

/**
 * Creates a debounced version of the provided function.
 * The debounced function delays invoking func until after
 * `delay` milliseconds have elapsed since the last invocation.
 *
 * @param func - The function to debounce
 * @param delay - Milliseconds to delay
 * @returns Debounced function with cancel method
 */
export function debounce<T extends (...args: unknown[]) => unknown>(
  func: T,
  delay: number
): {
  (...args: Parameters<T>): void;
  cancel: () => void;
  flush: () => void;
} {
  let timeoutId: ReturnType<typeof setTimeout> | null = null;
  let lastArgs: Parameters<T> | null = null;
  const debounced = (...args: Parameters<T>): void => {
    lastArgs = args;
    if (timeoutId !== null) {
      clearTimeout(timeoutId);
    }
    timeoutId = setTimeout(() => {
      timeoutId = null;
      lastArgs = null;
      func(...args);
    }, delay);
  };
  debounced.cancel = (): void => {
    if (timeoutId !== null) {
      clearTimeout(timeoutId);
      timeoutId = null;
      lastArgs = null;
    }
  };
  debounced.flush = (): void => {
    if (timeoutId !== null && lastArgs !== null) {
      clearTimeout(timeoutId);
      timeoutId = null;
      const args = lastArgs;
      lastArgs = null;
      func(...args);
    }
  };
  return debounced;
}

Copy the output with Cmd+C (Ghostty’s clipboard integration handles this cleanly) and paste into your editor pane.

# Scrollback works perfectly even for 1000+ line outputs
# Use Cmd+K to clear the screen between prompts
# Use Cmd+Shift+C to open a new tab for parallel Claude sessions

Common Issues

Text rendering artifacts after long sessions: Run Cmd+K to clear the screen, or restart the Ghostty tab. This is rare but can happen with 100K+ line scrollback.
Claude Code output truncated: Increase scrollback-limit in your config to 500000 if you regularly generate very long outputs.
Shell integration not working: Ensure shell-integration = zsh matches your actual shell. For bash, use shell-integration = bash.

Why This Matters

Ghostty’s GPU rendering means zero frame drops when Claude Code streams 500+ line responses. Developers report 30% less fatigue in extended sessions compared to CPU-rendered terminals.

Set it up → Build your permission config with our Permission Configurator.

Try it: Paste your error into our Error Diagnostic for an instant fix.

Frequently Asked Questions

What is the minimum setup required?

You need Claude Code installed (Node.js 18+), a project with a CLAUDE.md file, and the relevant toolchain for your project type (e.g., npm for JavaScript, pip for Python). The CLAUDE.md file should describe your project structure, conventions, and common commands so Claude Code can work effectively.

How long does the initial setup take?

For a typical project, initial setup takes 10-20 minutes. This includes creating the CLAUDE.md file, configuring .claude/settings.json for permissions, and running a test task to verify everything works. Subsequent sessions start immediately because the configuration persists.

Can I use this with a team?

Yes. Commit your .claude/ directory and CLAUDE.md to version control so the entire team uses the same configuration. Each developer can add personal preferences in ~/.claude/settings.json (user-level) without affecting the project configuration. Review CLAUDE.md changes in pull requests like any other configuration file.

What if Claude Code produces incorrect output?

First check that your CLAUDE.md accurately describes your project conventions. Incorrect or outdated context is the most common cause of wrong output. If the output is still wrong, provide feedback in the same session — Claude Code learns from corrections within a conversation. For persistent issues, add explicit rules to CLAUDE.md (e.g., “Always use single quotes” or “Never modify files in the config/ directory”).

Best Practices

Start with a clear CLAUDE.md. Describe your project structure, tech stack, coding conventions, and common commands in under 300 words. This single file has the largest impact on Claude Code’s accuracy and efficiency.
Use skills for domain knowledge. Move detailed reference information (API routes, database schemas, deployment procedures) into .claude/skills/ files. This keeps CLAUDE.md concise while making specialized knowledge available when needed.
Review changes before committing. Always run git diff after Claude Code makes changes. Verify the edits are correct, match your project style, and do not introduce unintended side effects. This habit prevents compounding errors across sessions.
Set up permission guardrails. Configure .claude/settings.json with explicit allow and deny lists. Allow your standard development commands (test, build, lint) and deny destructive operations (rm -rf, git push –force, database drops).
Keep sessions focused. Give Claude Code one clear task per prompt. Multi-step requests like “refactor auth, add tests, and update docs” produce better results when broken into three separate prompts, each building on the previous result.