feat: add token tracking and ASCII art interface

Add comprehensive token usage tracking and beautiful ASCII art interface to enhance user experience and provide cost transparency. - Add TokenTracker class for real-time token usage monitoring - Implement ASCII art banner with pyfiglet integration - Add cost estimation based on OpenAI pricing models - Display real-time token usage after each AI operation - Add 'tokens' command in interactive mode for usage statistics - Show session summary with detailed analytics at session end - Integrate token tracking into OpenAIProvider with optional TokenTracker - Add pyfiglet and tiktoken dependencies to pyproject.toml - Create ascii_art.py module for banner display functionality - Update main.py to initialize and manage token tracking sessions - Add session management with unique session IDs - Implement detailed usage tables with rich formatting - Update README.md with token tracking features and examples - Add ASCII art banner to welcome screen documentation - Include cost transparency and usage analytics sections - Update version to 0.2.5 in README examples - Remove Windows from supported OS list in bug report template - Add comprehensive unit tests for TokenTracker functionality - Add unit tests for ASCII art display features - Ensure proper error handling and edge cases covered - Real-time cost estimates displayed after each operation - Beautiful ASCII art welcome banner that adapts to terminal width - Session-based token tracking with detailed summaries - Interactive 'tokens' command for on-demand usage statistics - Rich formatted tables for better readability Breaking Changes: None Dependencies Added: pyfiglet>=1.0.0, tiktoken>=0.5.0
2025-10-09 13:42:56 +03:00 · 2025-08-22 11:30:11 +05:30
parent 48b2d8482e
commit 5f54f63b13
10 changed files with 1233 additions and 8 deletions
--- a/.github/ISSUE_TEMPLATE/bug_report.md
+++ b/.github/ISSUE_TEMPLATE/bug_report.md
@@ -15,7 +15,7 @@ A clear and concise description of what the bug is.
 3. Expected behavior: `...`

 ## Environment
- **OS**: [e.g., macOS 14.0, Ubuntu 22.04, Windows 11]
+- **OS**: [e.g., macOS 14.0, Ubuntu 22.04]
 - **Python version**: [e.g., 3.10.0]
 - **aiclip version**: [e.g., 0.2.2]
 - **ffmpeg version**: [e.g., 6.0]
--- a/README.md
+++ b/README.md
@@ -20,6 +20,8 @@
 - ⚡ **10x Faster**: Skip the documentation, Stack Overflow, and trial-and-error
 - 🎯 **Battle-Tested**: Generates reliable, production-ready commands
 - 🔄 **Smart Defaults**: Sensible codec and quality settings out of the box
+- 🎨 **Beautiful Interface**: Colorful ASCII art and responsive terminal UI
+- 📊 **Token Tracking**: Monitor your AI usage and costs in real-time

 ```bash
 # Instead of this...
@@ -57,9 +59,19 @@ aiclip
 ```

 ```text
+                                       ___    ____________    ________ 
+                                      /   |  /  _/ ____/ /   /  _/ __ \
+                                     / /| |  / // /   / /    / // /_/ /
+                                    / ___ |_/ // /___/ /____/ // ____/ 
+                                   /_/  |_/___/\____/_____/___/_/      
+
+                                    AI-Powered Video & Audio Processing
+
+                       ────────────────────────────────────────────────────────────
+
 ╭─────────────────────────────────────── Welcome to Interactive Mode ───────────────────────────────────────╮
 │                                                                                                           │
-│  ai-ffmpeg-cli v0.2.2                                                                                     │
+│  ai-ffmpeg-cli v0.2.5                                                                                     │
 │                                                                                                           │
 │  AI-powered video and audio processing with natural language                                              │
 │  Type your request in plain English and let AI handle the ffmpeg complexity!                              │
@@ -82,6 +94,8 @@ aiclip

 aiclip> convert this video to 720p

+📊 parse_intent: 2,073 → 47 (2,120 total) | $0.0057
+
 ┏━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━┓
 ┃ # ┃ Command                                      ┃ Output                                        ┃ Status ┃
 ┡━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━┩
@@ -186,6 +200,17 @@ aiclip --verbose "your command"
 aiclip --output-dir /path/to/output "convert video.mp4 to 720p"
 ```

+### Interactive Mode Features
+
+```bash
+# View token usage statistics anytime
+aiclip> tokens
+
+# Beautiful ASCII art welcome screen
+# Responsive terminal UI that adapts to your screen width
+# Real-time token usage tracking with cost estimates
+```
+
 ### Subcommands and option placement

 You can also use the explicit `nl` subcommand. Put global options before the subcommand:
@@ -225,6 +250,8 @@ AICLIP_OUTPUT_DIR=aiclip        # Default output directory
 - **Context Aware**: Scans your directory to suggest input files and durations
 - **Organized Output**: All generated files are saved to a dedicated output directory
 - **Duration Support**: Automatically handles time-based requests (e.g., "5 second GIF")
+- **Token Transparency**: Real-time tracking of AI usage and costs
+- **Responsive UI**: ASCII art and tables adapt to your terminal width

 ## 📁 Output Directory Management

@@ -251,6 +278,41 @@ aiclip "convert video.mp4 to 720p"
 - 🧹 **Clean workspace**: Input files stay separate from outputs
 - 📊 **Progress tracking**: See all your generated files at a glance

+## 📊 Token Usage Tracking
+
+aiclip provides complete transparency into your AI usage and costs:
+
+```bash
+# Real-time token usage displayed after each operation
+aiclip> convert video.mp4 to 720p
+📊 parse_intent: 2,073 → 47 (2,120 total) | $0.0057
+
+# View detailed session statistics anytime
+aiclip> tokens
+
+                     Token Usage Summary                     
+┏━━━━━━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
+┃ Metric         ┃    Value ┃ Details                       ┃
+┡━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
+│ Session ID     ┃ a1b2c3d4 ┃ Started at 14:30:15           │
+│ Duration       ┃     45.2s ┃ Session duration              │
+│ Operations     ┃        3 ┃ Total operations performed    │
+│ Input Tokens   ┃    6,142 ┃ Total tokens sent to AI       │
+│ Output Tokens  ┃      156 ┃ Total tokens received from AI │
+│ Total Tokens   ┃    6,298 ┃ Combined input + output       │
+│ Estimated Cost ┃  $0.0171 ┃ Based on OpenAI pricing       │
+└────────────────┴──────────┴───────────────────────────────┘
+
+# Session summary displayed at the end of each session
+```
+
+**Features:**
+- 💰 **Cost Transparency**: Real-time cost estimates based on current OpenAI pricing
+- 📈 **Usage Analytics**: Track tokens per operation and session totals
+- 🕒 **Session Tracking**: Monitor duration and operation counts
+- 🎯 **Model-Specific**: Accurate token counting for GPT-4o, GPT-3.5-turbo, and more
+- 📊 **Beautiful Tables**: Rich formatting that adapts to your terminal
+
 ## ⏱️ Duration and Time Handling

 aiclip intelligently handles time-based requests for video and GIF creation:
@@ -314,6 +376,7 @@ make demo
  - Ubuntu: `sudo apt install ffmpeg`
  - Windows: Download from [ffmpeg.org](https://ffmpeg.org/)
 - **OpenAI API key** for natural language processing
+- **Terminal with color support** for the best visual experience

 ## 🆘 Troubleshooting

@@ -378,6 +441,8 @@ See our [Contributing Guide](CONTRIBUTING.md) to get started.
 - 🔌 **Integrations**: GitHub Actions, Docker, CI/CD pipelines
 - 🎬 **Enhanced Duration Support**: Better handling of time-based requests
 - 📁 **Advanced Output Management**: Custom naming patterns and organization
+- 📊 **Usage Analytics**: Historical token usage tracking and reporting
+- 🎨 **Theme Customization**: Customizable color schemes and ASCII art fonts

 ## 📄 License

--- a/pyproject.toml
+++ b/pyproject.toml
@@ -52,7 +52,9 @@ dependencies = [
    "openai>=1.37.0",
    "python-dotenv>=1.0.0", 
    "pydantic>=2.0.0",
-    "typing-extensions>=4.8.0"
+    "typing-extensions>=4.8.0",
+    "pyfiglet>=1.0.0",
+    "tiktoken>=0.5.0"
 ]

 [project.optional-dependencies]
--- a/src/ai_ffmpeg_cli/ascii_art.py
+++ b/src/ai_ffmpeg_cli/ascii_art.py
@@ -0,0 +1,355 @@
+"""ASCII art module for ai-ffmpeg-cli.
+
+This module provides colorful ASCII art displays using pyfiglet
+with colors relevant to video and audio processing.
+"""
+
+from __future__ import annotations
+
+import random
+import shutil
+
+import pyfiglet
+from rich.console import Console
+from rich.text import Text
+
+# Initialize console for Rich output
+console = Console()
+
+
+def get_terminal_width() -> int:
+    """Get the current terminal width.
+
+    Returns:
+        Terminal width in characters, defaults to 80 if cannot be determined
+    """
+    try:
+        return shutil.get_terminal_size().columns
+    except (OSError, AttributeError):
+        return 80
+
+
+def center_text(text: str, width: int) -> str:
+    """Center-align text within a given width.
+
+    Args:
+        text: Text to center
+        width: Total width to center within
+
+    Returns:
+        Center-aligned text
+    """
+    lines = text.split("\n")
+    centered_lines = []
+
+    for line in lines:
+        if line.strip():
+            # Calculate padding to center the line
+            padding = max(0, (width - len(line)) // 2)
+            centered_line = " " * padding + line
+        else:
+            centered_line = line
+        centered_lines.append(centered_line)
+
+    return "\n".join(centered_lines)
+
+
+def get_ascii_art(text: str = "AICLIP", font: str = "slant") -> str:
+    """Generate ASCII art using pyfiglet.
+
+    Args:
+        text: Text to convert to ASCII art
+        font: Pyfiglet font to use
+
+    Returns:
+        ASCII art string
+    """
+    try:
+        fig = pyfiglet.Figlet(font=font)
+        ascii_art = fig.renderText(text)
+
+        # Center the ASCII art based on terminal width
+        terminal_width = get_terminal_width()
+        centered_art = center_text(ascii_art, terminal_width)
+
+        return centered_art
+    except Exception:
+        # Fallback to a simple ASCII art if pyfiglet fails
+        fallback_art = f"""
+╔══════════════════════════════════════════════════════════════╗
+║                        {text}                                ║
+╚══════════════════════════════════════════════════════════════╝
+"""
+        terminal_width = get_terminal_width()
+        return center_text(fallback_art, terminal_width)
+
+
+def display_colored_ascii_art(
+    text: str = "AICLIP",
+    font: str = "slant",
+    primary_color: str = "cyan",
+    secondary_color: str = "blue",
+    accent_color: str = "green",
+) -> None:
+    """Display colorful ASCII art with video/audio processing theme colors.
+
+    Args:
+        text: Text to display as ASCII art
+        font: Pyfiglet font to use
+        primary_color: Main color for the text (cyan - represents video/technology)
+        secondary_color: Secondary color for accents (blue - represents audio/water)
+        accent_color: Accent color for highlights (green - represents success/processing)
+    """
+    ascii_art = get_ascii_art(text, font)
+
+    # Create colored text with gradient effect
+    colored_text = Text()
+
+    # Split the ASCII art into lines
+    lines = ascii_art.split("\n")
+
+    for i, line in enumerate(lines):
+        if not line.strip():
+            colored_text.append("\n")
+            continue
+
+        # Create a gradient effect using different colors
+        for j, char in enumerate(line):
+            if char.isspace():
+                colored_text.append(char)
+            else:
+                # Alternate between primary and secondary colors for gradient effect
+                if (i + j) % 3 == 0:
+                    colored_text.append(char, style=f"bold {primary_color}")
+                elif (i + j) % 3 == 1:
+                    colored_text.append(char, style=f"bold {secondary_color}")
+                else:
+                    colored_text.append(char, style=f"bold {accent_color}")
+
+        colored_text.append("\n")
+
+    # Display the colored ASCII art
+    console.print(colored_text)
+
+    # Add a subtitle with accent color, center-aligned
+    subtitle = Text()
+    subtitle.append("AI-Powered Video & Audio Processing", style=f"italic {accent_color}")
+    console.print(subtitle, justify="center")
+    console.print()
+
+
+def display_welcome_banner() -> None:
+    """Display a welcome banner with ASCII art and project information."""
+    # Display the main ASCII art
+    display_colored_ascii_art()
+
+    # Add a decorative line, center-aligned
+    terminal_width = get_terminal_width()
+    line_length = min(60, terminal_width - 4)  # Leave some margin
+    padding = max(0, (terminal_width - line_length) // 2)
+    decorative_line = " " * padding + "─" * line_length
+    console.print(decorative_line, style="dim cyan")
+    console.print()
+
+
+def get_random_font() -> str:
+    """Get a random pyfiglet font for variety.
+
+    Returns:
+        Random font name
+    """
+    fonts = [
+        "slant",
+        "banner",
+        "big",
+        "block",
+        "bubble",
+        "digital",
+        "isometric1",
+        "isometric2",
+        "isometric3",
+        "isometric4",
+        "letters",
+        "alligator",
+        "dotmatrix",
+        "bubblehead",
+        "bulbhead",
+        "chunky",
+        "coinstak",
+        "colossal",
+        "crawford",
+        "diamond",
+        "epic",
+        "fender",
+        "fourtops",
+        "goofy",
+        "graceful",
+        "hollywood",
+        "invita",
+        "isometric1",
+        "isometric2",
+        "isometric3",
+        "isometric4",
+        "italic",
+        "larry3d",
+        "lcd",
+        "lean",
+        "letters",
+        "alligator2",
+        "alligator3",
+        "alphabet",
+        "arrows",
+        "avatar",
+        "banner3-D",
+        "banner3",
+        "banner4",
+        "barbwire",
+        "basic",
+        "bell",
+        "bigchief",
+        "binary",
+        "block",
+        "bubble",
+        "caligraphy",
+        "caligraphy2",
+        "catwalk",
+        "chunky",
+        "coinstak",
+        "colossal",
+        "computer",
+        "contessa",
+        "contrast",
+        "cosmic",
+        "cosmike",
+        "crawford",
+        "crawford2",
+        "crazy",
+        "cricket",
+        "cyberlarge",
+        "cybermedium",
+        "cybersmall",
+        "diamond",
+        "digital",
+        "doh",
+        "doom",
+        "dotmatrix",
+        "drpepper",
+        "eftichess",
+        "eftifont",
+        "eftipiti",
+        "eftirobot",
+        "eftitalic",
+        "eftiwall",
+        "eftiwater",
+        "epic",
+        "fender",
+        "fourtops",
+        "fuzzy",
+        "goofy",
+        "gothic",
+        "graceful",
+        "gradient",
+        "graffiti",
+        "hollywood",
+        "invita",
+        "isometric1",
+        "isometric2",
+        "isometric3",
+        "isometric4",
+        "italic",
+        "ivrit",
+        "jacky",
+        "katakana",
+        "kban",
+        "larry3d",
+        "lcd",
+        "lean",
+        "letters",
+        "linux",
+        "lockergnome",
+        "madrid",
+        "marquee",
+        "maxfour",
+        "mike",
+        "mini",
+        "mirror",
+        "mnemonic",
+        "morse",
+        "moscow",
+        "nancyj",
+        "nancyj-fancy",
+        "nancyj-underlined",
+        "nipples",
+        "ntgreek",
+        "o8",
+        "ogre",
+        "pawp",
+        "peaks",
+        "pebbles",
+        "pepper",
+        "poison",
+        "puffy",
+        "pyramid",
+        "rectangles",
+        "relief",
+        "relief2",
+        "rev",
+        "roman",
+        "rot13",
+        "rounded",
+        "rowancap",
+        "rozzo",
+        "runic",
+        "runyc",
+        "sblood",
+        "script",
+        "serifcap",
+        "shadow",
+        "short",
+        "slant",
+        "slide",
+        "slscript",
+        "small",
+        "smisome1",
+        "smkeyboard",
+        "smscript",
+        "smshadow",
+        "smslant",
+        "smtengwar",
+        "speed",
+        "stampatello",
+        "standard",
+        "starwars",
+        "stellar",
+        "stop",
+        "straight",
+        "tanja",
+        "tengwar",
+        "term",
+        "thick",
+        "thin",
+        "threepoint",
+        "ticks",
+        "ticksslant",
+        "tiles",
+        "tinker-toy",
+        "tombstone",
+        "trek",
+        "tsalagi",
+        "twopoint",
+        "univers",
+        "usaflag",
+        "wavy",
+        "weird",
+    ]
+
+    return random.choice(fonts)
+
+
+def display_dynamic_banner() -> None:
+    """Display a dynamic banner with random font and colors."""
+    # Choose a random font for variety
+    font = get_random_font()
+
+    # Display with the chosen font
+    display_colored_ascii_art(font=font)
--- a/src/ai_ffmpeg_cli/llm_client.py
+++ b/src/ai_ffmpeg_cli/llm_client.py
@@ -7,6 +7,7 @@ to parse natural language prompts into structured ffmpeg intents.
 from __future__ import annotations

 import json
+from typing import TYPE_CHECKING
 from typing import Any

 from pydantic import ValidationError
@@ -16,6 +17,9 @@ from .credential_security import sanitize_error_message
 from .custom_exceptions import ParseError
 from .intent_models import FfmpegIntent

+if TYPE_CHECKING:
+    from .token_tracker import TokenTracker
+
 # Create secure logger that masks sensitive information
 logger = create_secure_logger(__name__)

@@ -155,12 +159,13 @@ class OpenAIProvider(LLMProvider):
    including error handling and response processing.
    """

-    def __init__(self, api_key: str, model: str) -> None:
+    def __init__(self, api_key: str, model: str, token_tracker: TokenTracker | None = None) -> None:
        """Initialize OpenAI provider with API key and model.

        Args:
            api_key: OpenAI API key for authentication
            model: Model name to use for completions
+            token_tracker: Optional token tracker for monitoring usage

        Raises:
            Exception: When client initialization fails
@@ -173,6 +178,7 @@ class OpenAIProvider(LLMProvider):
        try:
            self.client = OpenAI(api_key=api_key)
            self.model = model
+            self.token_tracker = token_tracker
        except Exception as e:
            # Sanitize error message to prevent API key exposure
            sanitized_error = sanitize_error_message(str(e))
@@ -211,6 +217,30 @@ class OpenAIProvider(LLMProvider):

            content = rsp.choices[0].message.content or "{}"
            logger.debug(f"Received response length: {len(content)} characters")
+
+            # Track token usage if token tracker is available
+            if self.token_tracker and hasattr(rsp, "usage") and rsp.usage is not None:
+                input_tokens = rsp.usage.prompt_tokens if hasattr(rsp.usage, "prompt_tokens") else 0
+                output_tokens = (
+                    rsp.usage.completion_tokens if hasattr(rsp.usage, "completion_tokens") else 0
+                )
+                # Calculate cost estimate
+                cost_estimate = self.token_tracker.get_cost_estimate(
+                    self.model, input_tokens, output_tokens
+                )
+
+                # Track the operation
+                operation = self.token_tracker.track_operation(
+                    operation="parse_intent",
+                    model=self.model,
+                    input_text=system + "\n" + user,
+                    output_text=content,
+                    cost_estimate=cost_estimate,
+                )
+
+                # Display real-time usage
+                self.token_tracker.display_realtime_usage(operation)
+
            return content

        except Exception as e:
--- a/src/ai_ffmpeg_cli/logging_config/init.py
+++ b/src/ai_ffmpeg_cli/logging_config/init.py
@@ -28,4 +28,4 @@ __all__ = [
    "tenant_id",
 ]

-__version__ = "0.2.3"
+__version__ = "0.2.5"
--- a/src/ai_ffmpeg_cli/main.py
+++ b/src/ai_ffmpeg_cli/main.py
@@ -18,6 +18,7 @@ from rich.panel import Panel
 from rich.table import Table
 from rich.text import Text

+from .ascii_art import display_welcome_banner
 from .command_builder import build_commands
 from .config import AppConfig
 from .config import load_config
@@ -30,6 +31,7 @@ from .custom_exceptions import ParseError
 from .intent_router import route_intent
 from .llm_client import LLMClient
 from .llm_client import OpenAIProvider
+from .token_tracker import TokenTracker
 from .version_info import __version__

 # Initialize console for Rich output
@@ -38,13 +40,19 @@ console = Console()
 # Initialize logger
 logger = logging.getLogger(__name__)

+# Initialize global token tracker
+token_tracker = TokenTracker()
+
 # Initialize Typer app with completion disabled and support for invocation without subcommands
 app = typer.Typer(add_completion=False, help="AI-powered ffmpeg CLI", invoke_without_command=True)


 def _display_welcome_screen() -> None:
    """Display a beautiful welcome screen for the interactive mode."""
-    # Create welcome panel
+    # Display ASCII art banner
+    display_welcome_banner()
+
+    # Create welcome panel with version info
    welcome_text = Text()
    welcome_text.append("ai-ffmpeg-cli", style="bold white")
    welcome_text.append(" v", style="dim")
@@ -239,6 +247,7 @@ def _display_help_tips() -> None:
        "Try: 'add subtitles to video.mp4'",
        "Type 'exit' or 'quit' to leave interactive mode",
        "Use Ctrl+C to cancel any operation",
+        "Type 'tokens' to see usage statistics",
    ]

    tip_text = Text()
@@ -259,6 +268,23 @@ def _display_help_tips() -> None:
    console.print()


+def _start_token_session(cfg: AppConfig) -> None:
+    """Start a new token tracking session."""
+    import uuid
+
+    session_id = str(uuid.uuid4())[:8]  # Short session ID
+    token_tracker.start_session(session_id, cfg.model)
+    logger.debug(f"Started token tracking session: {session_id}")
+
+
+def _display_token_summary() -> None:
+    """Display token usage summary at the end of session."""
+    if token_tracker.current_session and token_tracker.current_session.operations:
+        console.print()
+        token_tracker.display_session_summary()
+        token_tracker.display_detailed_usage()
+
+
 def _setup_logging(verbose: bool) -> None:
    """Configure logging based on verbosity level.

@@ -326,6 +352,9 @@ def _main_impl(
        if invoked_none:
            if prompt is not None:
                try:
+                    # Start token tracking session for one-shot command
+                    _start_token_session(cfg)
+
                    # Execute one-shot command: scan context, parse intent, build and execute
                    context = scan(show_summary=False)  # Don't show summary for one-shot commands
                    client = _make_llm(cfg)
@@ -350,6 +379,10 @@ def _main_impl(
                        assume_yes=yes,
                        output_dir=Path(cfg.output_directory),
                    )
+
+                    # Display token summary for one-shot command
+                    _display_token_summary()
+
                    raise typer.Exit(code)
                except (ParseError, BuildError, ExecError) as e:
                    console.print(f"[red]❌ Error:[/red] {e}")
@@ -357,8 +390,12 @@ def _main_impl(
            else:
                # No subcommand and no prompt: enter interactive mode
                if ctx is not None:
+                    # Start token tracking session
+                    _start_token_session(cfg)
                    nl(ctx=ctx, prompt=None)
-                return
+                    # Display token summary at the end
+                    _display_token_summary()
+                    return
    except ConfigError as e:
        console.print(f"[red]❌ Configuration Error:[/red] {e}")
        raise typer.Exit(1) from e
@@ -432,7 +469,7 @@ def _make_llm(cfg: AppConfig) -> LLMClient:
    try:
        # This will validate the API key format and presence
        api_key = cfg.get_api_key_for_client()
-        provider = OpenAIProvider(api_key=api_key, model=cfg.model)
+        provider = OpenAIProvider(api_key=api_key, model=cfg.model, token_tracker=token_tracker)
        return LLMClient(provider)
    except ConfigError:
        # Re-raise config errors for proper error handling
@@ -532,6 +569,14 @@ def nl(
                if not line or line.lower() in {"exit", "quit"}:
                    console.print("[yellow]Goodbye![/yellow]")
                    break
+                if line.lower() == "tokens":
+                    # Display token usage statistics
+                    if token_tracker.current_session and token_tracker.current_session.operations:
+                        token_tracker.display_session_summary()
+                        token_tracker.display_detailed_usage()
+                    else:
+                        console.print("[dim]No token usage data available yet.[/dim]")
+                    continue
                try:
                    handle_one(line)
                except (ParseError, BuildError, ExecError) as e:
--- a/src/ai_ffmpeg_cli/token_tracker.py
+++ b/src/ai_ffmpeg_cli/token_tracker.py
@@ -0,0 +1,275 @@
+"""Token tracking module for ai-ffmpeg-cli.
+
+This module provides functionality to track token usage using tiktoken
+and display usage statistics to users.
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from dataclasses import field
+from datetime import datetime
+
+import tiktoken
+from rich.console import Console
+from rich.table import Table
+from rich.text import Text
+
+# Initialize console for Rich output
+console = Console()
+
+
+@dataclass
+class TokenUsage:
+    """Represents token usage for a single operation."""
+
+    operation: str
+    model: str
+    input_tokens: int
+    output_tokens: int
+    total_tokens: int
+    timestamp: datetime = field(default_factory=datetime.now)
+    cost_estimate: float | None = None
+
+    @property
+    def total_cost_estimate(self) -> float | None:
+        """Calculate total cost estimate if available."""
+        if self.cost_estimate:
+            return self.cost_estimate
+        return None
+
+
+@dataclass
+class SessionTokenTracker:
+    """Tracks token usage for a CLI session."""
+
+    session_id: str
+    start_time: datetime = field(default_factory=datetime.now)
+    operations: list[TokenUsage] = field(default_factory=list)
+    model: str = "gpt-4o"
+
+    def add_operation(self, operation: TokenUsage) -> None:
+        """Add a token usage operation to the session."""
+        self.operations.append(operation)
+
+    @property
+    def total_input_tokens(self) -> int:
+        """Get total input tokens for the session."""
+        return sum(op.input_tokens for op in self.operations)
+
+    @property
+    def total_output_tokens(self) -> int:
+        """Get total output tokens for the session."""
+        return sum(op.output_tokens for op in self.operations)
+
+    @property
+    def total_tokens(self) -> int:
+        """Get total tokens for the session."""
+        return sum(op.total_tokens for op in self.operations)
+
+    @property
+    def total_cost_estimate(self) -> float | None:
+        """Get total cost estimate for the session."""
+        costs = [op.total_cost_estimate for op in self.operations if op.total_cost_estimate]
+        return sum(costs) if costs else None
+
+    @property
+    def session_duration(self) -> float:
+        """Get session duration in seconds."""
+        return (datetime.now() - self.start_time).total_seconds()
+
+
+class TokenTracker:
+    """Main token tracking class."""
+
+    def __init__(self) -> None:
+        self.current_session: SessionTokenTracker | None = None
+        self.encoding_cache: dict[str, tiktoken.Encoding] = {}
+
+    def start_session(self, session_id: str, model: str = "gpt-4o") -> None:
+        """Start a new token tracking session."""
+        self.current_session = SessionTokenTracker(session_id=session_id, model=model)
+
+    def get_encoding(self, model: str) -> tiktoken.Encoding:
+        """Get tiktoken encoding for a model, with caching."""
+        if model not in self.encoding_cache:
+            try:
+                # Map common model names to their encoding
+                if "gpt-4" in model or "gpt-3.5" in model:
+                    encoding_name = "cl100k_base"  # GPT-4 and GPT-3.5-turbo
+                elif "gpt-3" in model:
+                    encoding_name = "r50k_base"  # GPT-3
+                else:
+                    encoding_name = "cl100k_base"  # Default to GPT-4 encoding
+
+                self.encoding_cache[model] = tiktoken.get_encoding(encoding_name)
+            except Exception:
+                # Fallback to GPT-4 encoding if model-specific encoding fails
+                self.encoding_cache[model] = tiktoken.get_encoding("cl100k_base")
+
+        return self.encoding_cache[model]
+
+    def count_tokens(self, text: str, model: str) -> int:
+        """Count tokens in text for a specific model."""
+        if not text:
+            return 0
+
+        encoding = self.get_encoding(model)
+        return len(encoding.encode(text))
+
+    def track_operation(
+        self,
+        operation: str,
+        model: str,
+        input_text: str,
+        output_text: str,
+        cost_estimate: float | None = None,
+    ) -> TokenUsage:
+        """Track token usage for an operation."""
+        if not self.current_session:
+            # Create a default session if none exists
+            self.start_session("default", model)
+
+        input_tokens = self.count_tokens(input_text, model)
+        output_tokens = self.count_tokens(output_text, model)
+        total_tokens = input_tokens + output_tokens
+
+        usage = TokenUsage(
+            operation=operation,
+            model=model,
+            input_tokens=input_tokens,
+            output_tokens=output_tokens,
+            total_tokens=total_tokens,
+            cost_estimate=cost_estimate,
+        )
+
+        if self.current_session:
+            self.current_session.add_operation(usage)
+        return usage
+
+    def display_session_summary(self) -> None:
+        """Display a summary of token usage for the current session."""
+        if not self.current_session or not self.current_session.operations:
+            return
+
+        session = self.current_session
+
+        # Create summary table
+        summary_table = Table(
+            title="[bold cyan]Token Usage Summary[/bold cyan]",
+            show_header=True,
+            header_style="bold blue",
+        )
+        summary_table.add_column("Metric", style="bold cyan")
+        summary_table.add_column("Value", style="white", justify="right")
+        summary_table.add_column("Details", style="dim")
+
+        # Session info
+        summary_table.add_row(
+            "Session ID",
+            session.session_id,
+            f"Started at {session.start_time.strftime('%H:%M:%S')}",
+        )
+
+        summary_table.add_row("Duration", f"{session.session_duration:.1f}s", "Session duration")
+
+        summary_table.add_row(
+            "Operations", str(len(session.operations)), "Total operations performed"
+        )
+
+        # Token counts
+        summary_table.add_row(
+            "Input Tokens", f"{session.total_input_tokens:,}", "Total tokens sent to AI"
+        )
+
+        summary_table.add_row(
+            "Output Tokens",
+            f"{session.total_output_tokens:,}",
+            "Total tokens received from AI",
+        )
+
+        summary_table.add_row(
+            "Total Tokens", f"{session.total_tokens:,}", "Combined input + output"
+        )
+
+        # Cost estimate if available
+        if session.total_cost_estimate:
+            summary_table.add_row(
+                "Estimated Cost",
+                f"${session.total_cost_estimate:.4f}",
+                "Based on OpenAI pricing",
+            )
+
+        console.print(summary_table)
+        console.print()
+
+    def display_detailed_usage(self) -> None:
+        """Display detailed token usage for each operation."""
+        if not self.current_session or not self.current_session.operations:
+            return
+
+        # Create detailed operations table
+        operations_table = Table(
+            title="[bold green]Detailed Token Usage[/bold green]",
+            show_header=True,
+            header_style="bold blue",
+        )
+        operations_table.add_column("#", style="bold cyan", justify="center")
+        operations_table.add_column("Operation", style="bold white")
+        operations_table.add_column("Model", style="cyan")
+        operations_table.add_column("Input", style="green", justify="right")
+        operations_table.add_column("Output", style="yellow", justify="right")
+        operations_table.add_column("Total", style="bold", justify="right")
+        operations_table.add_column("Time", style="dim")
+
+        for i, operation in enumerate(self.current_session.operations, 1):
+            operations_table.add_row(
+                str(i),
+                operation.operation,
+                operation.model,
+                f"{operation.input_tokens:,}",
+                f"{operation.output_tokens:,}",
+                f"{operation.total_tokens:,}",
+                operation.timestamp.strftime("%H:%M:%S"),
+            )
+
+        console.print(operations_table)
+        console.print()
+
+    def get_cost_estimate(self, model: str, input_tokens: int, output_tokens: int) -> float | None:
+        """Get cost estimate for token usage based on OpenAI pricing."""
+        # OpenAI pricing (as of 2024, approximate)
+        pricing = {
+            "gpt-4o": {"input": 0.0025, "output": 0.01},  # per 1K tokens
+            "gpt-4o-mini": {"input": 0.00015, "output": 0.0006},
+            "gpt-4-turbo": {"input": 0.01, "output": 0.03},
+            "gpt-3.5-turbo": {"input": 0.0005, "output": 0.0015},
+        }
+
+        # Find the closest model match
+        model_key = None
+        for key in pricing:
+            if key in model.lower():
+                model_key = key
+                break
+
+        if not model_key:
+            return None
+
+        input_cost = (input_tokens / 1000) * pricing[model_key]["input"]
+        output_cost = (output_tokens / 1000) * pricing[model_key]["output"]
+
+        return input_cost + output_cost
+
+    def display_realtime_usage(self, operation: TokenUsage) -> None:
+        """Display real-time token usage for an operation."""
+        usage_text = Text()
+        usage_text.append("📊 ", style="bold cyan")
+        usage_text.append(f"{operation.operation}: ", style="bold white")
+        usage_text.append(f"{operation.input_tokens:,} → {operation.output_tokens:,} ", style="dim")
+        usage_text.append(f"({operation.total_tokens:,} total)", style="green")
+
+        if operation.total_cost_estimate:
+            usage_text.append(f" | ${operation.total_cost_estimate:.4f}", style="yellow")
+
+        console.print(usage_text)
--- a/tests/unit/test_ascii_art.py
+++ b/tests/unit/test_ascii_art.py
@@ -0,0 +1,138 @@
+"""Tests for ASCII art functionality."""
+
+from unittest.mock import patch
+
+from ai_ffmpeg_cli.ascii_art import center_text
+from ai_ffmpeg_cli.ascii_art import display_colored_ascii_art
+from ai_ffmpeg_cli.ascii_art import display_dynamic_banner
+from ai_ffmpeg_cli.ascii_art import display_welcome_banner
+from ai_ffmpeg_cli.ascii_art import get_ascii_art
+from ai_ffmpeg_cli.ascii_art import get_random_font
+from ai_ffmpeg_cli.ascii_art import get_terminal_width
+
+
+class TestAsciiArt:
+    """Test cases for ASCII art functionality."""
+
+    def test_get_ascii_art_default(self):
+        """Test getting ASCII art with default parameters."""
+        result = get_ascii_art()
+        # Check that it's a non-empty string with ASCII art characters
+        assert isinstance(result, str)
+        assert len(result) > 0
+        # Should contain some ASCII art characters
+        assert any(char in result for char in ["/", "\\", "_", "|", "#", "="])
+
+    def test_get_ascii_art_custom_text(self):
+        """Test getting ASCII art with custom text."""
+        result = get_ascii_art("TEST")
+        assert isinstance(result, str)
+        assert len(result) > 0
+        # Should contain some ASCII art characters
+        assert any(char in result for char in ["/", "\\", "_", "|", "#", "="])
+
+    def test_get_ascii_art_custom_font(self):
+        """Test getting ASCII art with custom font."""
+        result = get_ascii_art("TEST", "banner")
+        assert isinstance(result, str)
+        assert len(result) > 0
+        # Should contain some ASCII art characters
+        assert any(char in result for char in ["/", "\\", "_", "|", "#", "="])
+
+    def test_get_ascii_art_fallback(self):
+        """Test ASCII art fallback when pyfiglet fails."""
+        with patch("ai_ffmpeg_cli.ascii_art.pyfiglet.Figlet") as mock_figlet:
+            mock_figlet.side_effect = Exception("Font not found")
+            result = get_ascii_art("TEST")
+            assert "TEST" in result
+            assert "╔" in result  # Check for fallback box characters
+
+    def test_get_random_font(self):
+        """Test getting a random font."""
+        font = get_random_font()
+        assert isinstance(font, str)
+        assert len(font) > 0
+
+    def test_get_random_font_returns_valid_font(self):
+        """Test that random font is from the valid font list."""
+        font = get_random_font()
+        # Since it's random, we can't guarantee it's in our small test list
+        # But we can test that it's a string and not empty
+        assert isinstance(font, str)
+        assert len(font) > 0
+
+    @patch("ai_ffmpeg_cli.ascii_art.console")
+    def test_display_colored_ascii_art(self, mock_console):
+        """Test displaying colored ASCII art."""
+        display_colored_ascii_art("TEST")
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    @patch("ai_ffmpeg_cli.ascii_art.console")
+    def test_display_colored_ascii_art_custom_colors(self, mock_console):
+        """Test displaying colored ASCII art with custom colors."""
+        display_colored_ascii_art(
+            "TEST", primary_color="red", secondary_color="yellow", accent_color="blue"
+        )
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    @patch("ai_ffmpeg_cli.ascii_art.console")
+    def test_display_welcome_banner(self, mock_console):
+        """Test displaying welcome banner."""
+        display_welcome_banner()
+        # Verify that console.print was called multiple times
+        assert mock_console.print.call_count >= 2
+
+    @patch("ai_ffmpeg_cli.ascii_art.console")
+    def test_display_dynamic_banner(self, mock_console):
+        """Test displaying dynamic banner."""
+        display_dynamic_banner()
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    def test_ascii_art_with_empty_text(self):
+        """Test ASCII art with empty text."""
+        result = get_ascii_art("")
+        assert isinstance(result, str)
+        # Should still produce some output even with empty text
+
+    def test_ascii_art_with_special_characters(self):
+        """Test ASCII art with special characters."""
+        result = get_ascii_art("AICLIP-123!")
+        assert isinstance(result, str)
+        assert len(result) > 0
+
+    @patch("ai_ffmpeg_cli.ascii_art.console")
+    def test_display_colored_ascii_art_with_long_text(self, mock_console):
+        """Test displaying colored ASCII art with long text."""
+        long_text = "VERY_LONG_TEXT_FOR_TESTING"
+        display_colored_ascii_art(long_text)
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    def test_get_terminal_width(self):
+        """Test getting terminal width."""
+        width = get_terminal_width()
+        assert isinstance(width, int)
+        assert width > 0
+
+    def test_center_text(self):
+        """Test centering text."""
+        text = "Hello\nWorld"
+        centered = center_text(text, 20)
+        assert isinstance(centered, str)
+        assert "Hello" in centered
+        assert "World" in centered
+
+    def test_center_text_empty(self):
+        """Test centering empty text."""
+        text = ""
+        centered = center_text(text, 20)
+        assert centered == ""
+
+    def test_center_text_short_width(self):
+        """Test centering text when width is shorter than text."""
+        text = "Hello World"
+        centered = center_text(text, 5)
+        assert "Hello World" in centered
--- a/tests/unit/test_token_tracker.py
+++ b/tests/unit/test_token_tracker.py
@@ -0,0 +1,315 @@
+"""Tests for token tracking functionality."""
+
+from datetime import datetime
+from unittest.mock import MagicMock
+from unittest.mock import patch
+
+from ai_ffmpeg_cli.token_tracker import SessionTokenTracker
+from ai_ffmpeg_cli.token_tracker import TokenTracker
+from ai_ffmpeg_cli.token_tracker import TokenUsage
+
+
+class TestTokenUsage:
+    """Test cases for TokenUsage class."""
+
+    def test_token_usage_creation(self):
+        """Test creating a TokenUsage instance."""
+        usage = TokenUsage(
+            operation="test_operation",
+            model="gpt-4o",
+            input_tokens=100,
+            output_tokens=50,
+            total_tokens=150,
+        )
+
+        assert usage.operation == "test_operation"
+        assert usage.model == "gpt-4o"
+        assert usage.input_tokens == 100
+        assert usage.output_tokens == 50
+        assert usage.total_tokens == 150
+        assert usage.cost_estimate is None
+
+    def test_token_usage_with_cost(self):
+        """Test TokenUsage with cost estimate."""
+        usage = TokenUsage(
+            operation="test_operation",
+            model="gpt-4o",
+            input_tokens=100,
+            output_tokens=50,
+            total_tokens=150,
+            cost_estimate=0.0025,
+        )
+
+        assert usage.total_cost_estimate == 0.0025
+
+    def test_token_usage_timestamp(self):
+        """Test that TokenUsage has a timestamp."""
+        usage = TokenUsage(
+            operation="test_operation",
+            model="gpt-4o",
+            input_tokens=100,
+            output_tokens=50,
+            total_tokens=150,
+        )
+
+        assert isinstance(usage.timestamp, datetime)
+
+
+class TestSessionTokenTracker:
+    """Test cases for SessionTokenTracker class."""
+
+    def test_session_creation(self):
+        """Test creating a session tracker."""
+        session = SessionTokenTracker(session_id="test123", model="gpt-4o")
+
+        assert session.session_id == "test123"
+        assert session.model == "gpt-4o"
+        assert len(session.operations) == 0
+        assert isinstance(session.start_time, datetime)
+
+    def test_add_operation(self):
+        """Test adding operations to session."""
+        session = SessionTokenTracker(session_id="test123")
+        usage = TokenUsage(
+            operation="test_op",
+            model="gpt-4o",
+            input_tokens=100,
+            output_tokens=50,
+            total_tokens=150,
+        )
+
+        session.add_operation(usage)
+        assert len(session.operations) == 1
+        assert session.operations[0] == usage
+
+    def test_total_tokens_calculation(self):
+        """Test total token calculations."""
+        session = SessionTokenTracker(session_id="test123")
+
+        # Add multiple operations
+        session.add_operation(TokenUsage("op1", "gpt-4o", 100, 50, 150))
+        session.add_operation(TokenUsage("op2", "gpt-4o", 200, 100, 300))
+
+        assert session.total_input_tokens == 300
+        assert session.total_output_tokens == 150
+        assert session.total_tokens == 450
+
+    def test_cost_calculation(self):
+        """Test cost calculation with multiple operations."""
+        session = SessionTokenTracker(session_id="test123")
+
+        # Add operations with costs
+        op1 = TokenUsage("op1", "gpt-4o", 100, 50, 150, cost_estimate=0.001)
+        op2 = TokenUsage("op2", "gpt-4o", 200, 100, 300, cost_estimate=0.002)
+        session.add_operation(op1)
+        session.add_operation(op2)
+
+        assert session.total_cost_estimate == 0.003
+
+    def test_session_duration(self):
+        """Test session duration calculation."""
+        session = SessionTokenTracker(session_id="test123")
+
+        # Duration should be a positive number
+        assert session.session_duration >= 0
+
+
+class TestTokenTracker:
+    """Test cases for TokenTracker class."""
+
+    def test_token_tracker_initialization(self):
+        """Test TokenTracker initialization."""
+        tracker = TokenTracker()
+
+        assert tracker.current_session is None
+        assert isinstance(tracker.encoding_cache, dict)
+        assert len(tracker.encoding_cache) == 0
+
+    def test_start_session(self):
+        """Test starting a new session."""
+        tracker = TokenTracker()
+        tracker.start_session("test123", "gpt-4o")
+
+        assert tracker.current_session is not None
+        assert tracker.current_session.session_id == "test123"
+        assert tracker.current_session.model == "gpt-4o"
+
+    @patch("ai_ffmpeg_cli.token_tracker.tiktoken.get_encoding")
+    def test_get_encoding_gpt4(self, mock_get_encoding):
+        """Test getting encoding for GPT-4 model."""
+        mock_encoding = MagicMock()
+        mock_get_encoding.return_value = mock_encoding
+
+        tracker = TokenTracker()
+        encoding = tracker.get_encoding("gpt-4o")
+
+        assert encoding == mock_encoding
+        mock_get_encoding.assert_called_with("cl100k_base")
+
+    @patch("ai_ffmpeg_cli.token_tracker.tiktoken.get_encoding")
+    def test_get_encoding_gpt3(self, mock_get_encoding):
+        """Test getting encoding for GPT-3 model."""
+        mock_encoding = MagicMock()
+        mock_get_encoding.return_value = mock_encoding
+
+        tracker = TokenTracker()
+        encoding = tracker.get_encoding("gpt-3")
+
+        assert encoding == mock_encoding
+        mock_get_encoding.assert_called_with("r50k_base")
+
+    @patch("ai_ffmpeg_cli.token_tracker.tiktoken.get_encoding")
+    def test_get_encoding_caching(self, mock_get_encoding):
+        """Test that encodings are cached."""
+        mock_encoding = MagicMock()
+        mock_get_encoding.return_value = mock_encoding
+
+        tracker = TokenTracker()
+
+        # First call
+        encoding1 = tracker.get_encoding("gpt-4o")
+        # Second call should use cache
+        encoding2 = tracker.get_encoding("gpt-4o")
+
+        assert encoding1 == encoding2
+        # Should only be called once due to caching
+        mock_get_encoding.assert_called_once_with("cl100k_base")
+
+    def test_count_tokens_empty(self):
+        """Test counting tokens in empty text."""
+        tracker = TokenTracker()
+        count = tracker.count_tokens("", "gpt-4o")
+        assert count == 0
+
+    def test_count_tokens_with_text(self):
+        """Test counting tokens in text."""
+        tracker = TokenTracker()
+        # Mock the encoding to return a known token count
+        mock_encoding = MagicMock()
+        mock_encoding.encode.return_value = [1, 2, 3, 4, 5]  # 5 tokens
+        tracker.encoding_cache["gpt-4o"] = mock_encoding
+
+        count = tracker.count_tokens("Hello world", "gpt-4o")
+        assert count == 5
+
+    def test_track_operation(self):
+        """Test tracking an operation."""
+        tracker = TokenTracker()
+        tracker.start_session("test123", "gpt-4o")
+
+        # Mock encoding for token counting
+        mock_encoding = MagicMock()
+        mock_encoding.encode.return_value = [1, 2, 3]  # 3 tokens for input
+        tracker.encoding_cache["gpt-4o"] = mock_encoding
+
+        usage = tracker.track_operation(
+            operation="test_op",
+            model="gpt-4o",
+            input_text="Hello",
+            output_text="World",
+            cost_estimate=0.001,
+        )
+
+        assert usage.operation == "test_op"
+        assert usage.model == "gpt-4o"
+        assert usage.input_tokens == 3
+        assert usage.output_tokens == 3  # "World" also has 3 tokens
+        assert usage.total_tokens == 6
+        assert usage.cost_estimate == 0.001
+
+        # Check that it was added to the session
+        assert len(tracker.current_session.operations) == 1
+
+    def test_track_operation_no_session(self):
+        """Test tracking operation when no session exists."""
+        tracker = TokenTracker()
+
+        # Mock encoding
+        mock_encoding = MagicMock()
+        mock_encoding.encode.return_value = [1, 2]
+        tracker.encoding_cache["gpt-4o"] = mock_encoding
+
+        tracker.track_operation(
+            operation="test_op", model="gpt-4o", input_text="Hello", output_text="World"
+        )
+
+        # Should create a default session
+        assert tracker.current_session is not None
+        assert tracker.current_session.session_id == "default"
+        assert len(tracker.current_session.operations) == 1
+
+    def test_get_cost_estimate_gpt4o(self):
+        """Test cost estimation for GPT-4o."""
+        tracker = TokenTracker()
+        cost = tracker.get_cost_estimate("gpt-4o", 1000, 500)
+
+        # GPT-4o pricing: input $0.0025/1K, output $0.01/1K
+        expected_cost = (1000 / 1000) * 0.0025 + (500 / 1000) * 0.01
+        assert cost == expected_cost
+
+    def test_get_cost_estimate_gpt35(self):
+        """Test cost estimation for GPT-3.5-turbo."""
+        tracker = TokenTracker()
+        cost = tracker.get_cost_estimate("gpt-3.5-turbo", 1000, 500)
+
+        # GPT-3.5-turbo pricing: input $0.0005/1K, output $0.0015/1K
+        expected_cost = (1000 / 1000) * 0.0005 + (500 / 1000) * 0.0015
+        assert cost == expected_cost
+
+    def test_get_cost_estimate_unknown_model(self):
+        """Test cost estimation for unknown model."""
+        tracker = TokenTracker()
+        cost = tracker.get_cost_estimate("unknown-model", 1000, 500)
+        assert cost is None
+
+    @patch("ai_ffmpeg_cli.token_tracker.console")
+    def test_display_session_summary(self, mock_console):
+        """Test displaying session summary."""
+        tracker = TokenTracker()
+        tracker.start_session("test123", "gpt-4o")
+
+        # Add some operations
+        tracker.track_operation("op1", "gpt-4o", "input1", "output1", 0.001)
+        tracker.track_operation("op2", "gpt-4o", "input2", "output2", 0.002)
+
+        tracker.display_session_summary()
+
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    @patch("ai_ffmpeg_cli.token_tracker.console")
+    def test_display_detailed_usage(self, mock_console):
+        """Test displaying detailed usage."""
+        tracker = TokenTracker()
+        tracker.start_session("test123", "gpt-4o")
+
+        # Add some operations
+        tracker.track_operation("op1", "gpt-4o", "input1", "output1")
+        tracker.track_operation("op2", "gpt-4o", "input2", "output2")
+
+        tracker.display_detailed_usage()
+
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+    @patch("ai_ffmpeg_cli.token_tracker.console")
+    def test_display_realtime_usage(self, mock_console):
+        """Test displaying real-time usage."""
+        tracker = TokenTracker()
+        usage = TokenUsage(
+            operation="test_op",
+            model="gpt-4o",
+            input_tokens=100,
+            output_tokens=50,
+            total_tokens=150,
+            cost_estimate=0.001,
+        )
+
+        tracker.display_realtime_usage(usage)
+
+        # Verify that console.print was called
+        assert mock_console.print.called
+
+
+# Note: get_terminal_width and center_text functions are tested in test_ascii_art.py
+# since they are part of the ascii_art module