Skip to main content

Crate yt_dlp

Crate yt_dlp 

Source
Expand description

🎬️ A Rust library (with auto dependencies installation) for video downloading

This library is a Rust asynchronous wrapper around the yt-dlp command line tool, a feature-rich audio/video downloader supporting YouTube, Vimeo, TikTok, Instagram, Twitter, and more.
The crate is designed to download audio and video from various websites. You don't need to care about dependencies, yt-dlp and ffmpeg will be downloaded automatically.

⚠️ The project is still in development, so if you encounter any bugs or have any feature requests, please open an issue or a discussion.


Develop CI Release Downloads Ask DeepWiki

Discussions Issues Pull%20requests

License Stars Forks

Statistics

Codecov Quality Scorecard


Β§πŸ’­οΈ Why use an external Python app?

Originally, to download videos from YouTube, I used the rustube crate, written in pure Rust and without any external dependencies. However, I quickly realized that due to frequent breaking changes on the YouTube website, the crate was outdated and no longer functional.

After a few tests and research, I concluded that the python app yt-dlp was the best compromise, thanks to its regular updates and massive community. Its standalone binaries and its ability to output the fetched data in JSON format make it a perfect candidate for a Rust wrapper.

Using an external program is not ideal, but it is the most reliable and maintained solution for now.

Β§πŸ“₯ How to get it

Add the following to your Cargo.toml file:

[dependencies]
yt-dlp = "2.7.0"

A new release is automatically published every two weeks, to keep up to date with dependencies and features. Make sure to check the releases page to see the latest version of the crate.

Β§πŸ”Œ Optional features

This library puts a lot of functionality behind optional features in order to optimize compile time for the most common use cases. The following features are available.

  • πŸͺ hooks - Enables Rust hooks and callbacks for download events. Allows registering async functions that will be called when events occur.
  • πŸ“‘ webhooks - Enables HTTP webhooks delivery for download events. Allows sending events to external HTTP endpoints with retry logic.
  • πŸ“Š statistics - Enables real-time statistics and analytics on downloads and fetches. Exposes aggregate counters, averages, success rates, and a bounded history window.
  • ⚑ cache-memory (enabled by default) β€” In-memory Moka cache (pulls in moka). Fast TTL-based eviction; no persistence.
  • πŸ—ƒοΈ cache-json β€” JSON file-system backend. One .json file per entry.
  • πŸ—„οΈ cache-redb β€” Embedded redb backend. Single-file, pure-rust, ACID-compliant.
  • 🌐 cache-redis β€” Distributed Redis backend. Native TTL via SETEX.
  • πŸ”΄ live-recording - Enables live stream recording via HLS segment fetching (reqwest) or FFmpeg fallback. Pulls in m3u8-rs for HLS manifest parsing.
  • πŸ“‘ live-streaming - Enables live fragment streaming via HLS segment fetching (reqwest). Pulls in m3u8-rs for HLS manifest parsing.
  • πŸ”’ rustls - Enables the rustls-tls feature in the reqwest crate. This enables building the application without openssl or other system sourced SSL libraries.
  • 🌍 hickory-dns - Enables async DNS resolution via Hickory DNS (passes reqwest/hickory-dns). Replaces the default blocking system resolver with a fully async, pure-Rust resolver.

Β§πŸ—„οΈ Cache backends

The library includes a tiered metadata cache that avoids redundant yt-dlp subprocess calls for video info, downloaded files, and playlists. The architecture uses an optional L1 in-memory layer (Moka) and an optional L2 persistent layer, selected exclusively via Cargo features:

FeatureBackendPersistenceNotes
cache-memory (default)In-memory Moka❌ NoTTL-based eviction, async-ready
cache-jsonJSON files on diskβœ… YesOne .json file per entry in the cache directory
cache-redbEmbedded redbβœ… YesSingle-file, pure-Rust, ACID transactions
cache-redisRedisβœ… YesDistributed, native TTL via SETEX

Multiple persistent backends can be compiled in simultaneously. When exactly one is enabled, it is selected automatically. When several are enabled, CacheConfig::persistent_backend must be set explicitly; otherwise CacheLayer::from_config returns an Error::AmbiguousCacheBackend at runtime. The cache-memory feature (Moka L1) can always be combined with any persistent backend for a tiered L1 + L2 setup.

Default (in-memory Moka) β€” no persistence, TTL-based eviction, useful for short-lived processes:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-memory"] }

JSON β€” persistent, file-system backed, no extra dependencies:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-json"] }

Redb β€” embedded, single-file, ACID-compliant, great for desktop/server apps:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-redb"] }

Redis β€” distributed, ideal for multi-node or cloud deployments:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-redis"] }

Tiered (Moka L1 + persistent L2) β€” best of both worlds:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-memory", "cache-redb"] }

Multiple backends compiled in β€” select one at runtime via CacheConfig::persistent_backend:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["cache-memory", "cache-json", "cache-redb"] }
β“˜
use yt_dlp::prelude::*;

let config = CacheConfig::builder()
    .cache_dir("cache")
    .persistent_backend(PersistentBackendKind::Redb) // required when multiple compiled in
    .build();
Β§CDN URL expiry and cache invalidation

When a Video is cached, its stream format URLs are valid for approximately 6 hours (YouTube CDN lifetime). The library tracks this automatically via the available_at field on each Format.

On every fetch_video_infos call, the cache checks whether the format URLs are still fresh. If they have expired, the cached entry is silently invalidated and the video is re-fetched β€” so you never end up downloading with stale CDN URLs. The configured TTL (default 24 h) acts as an upper bound; the effective TTL is min(configured_ttl, cdn_url_lifetime).

This behavior is transparent and requires no changes to your code. You can inspect the expiry yourself:

β“˜
if !video.are_format_urls_fresh() {
    // URLs are stale β€” fetch_video_infos will re-fetch automatically
}
// Or get the earliest available_at timestamp across all downloadable formats:
if let Some(ts) = video.formats_available_at() {
    println!("Format URLs valid until approx. {} (unix)", ts + yt_dlp::model::FORMAT_URL_LIFETIME);
}

Β§πŸ” Observability & Tracing

This crate always includes the Tracing tracing crate. The library emits debug and trace span events throughout its internal operations (downloads, cache lookups, subprocess execution, etc.).

⚠️ Important: tracing macros are pure no-ops without a configured subscriber. If you don’t add one, there is zero runtime overhead.

To capture logs, add a subscriber in your application:

[dependencies]
tracing-subscriber = "0.3"
β“˜
use tracing::Level;
use tracing_subscriber::FmtSubscriber;

let subscriber = FmtSubscriber::builder()
        // all spans/events with a level higher than TRACE (e.g, debug, info, warn, etc.)
        // will be written to stdout.
        .with_max_level(Level::TRACE)
        // completes the builder.
        .finish();
tracing::subscriber::set_global_default(subscriber)
        .expect("setting default subscriber failed");

Refer to the tracing-subscriber documentation for more advanced configuration (JSON output, log levels, targets, etc.).


Β§πŸ“– Documentation

The documentation is available on docs.rs.

Β§πŸ—οΈ Multi-Extractor Architecture

This library now supports downloading from 1,800+ websites through a flexible extractor system:

  • Downloader - Universal client supporting all sites via the extractors
  • extractor::Youtube - Highly optimized YouTube extractor with platform-specific features:
    • Player client selection (Android, iOS, Web, TvEmbedded) for bypassing restrictions
    • Format presets (Best, Premium, High, Medium, Low, AudioOnly, ModernCodecs)
    • YouTube-specific methods: search(), fetch_channel(), fetch_user(), fetch_playlist_paginated()
  • extractor::Generic - Universal extractor for all other sites with authentication support

§🧩 Usage Patterns

  • 🎬️ For YouTube with optimizations:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg")
    );
    let downloader = Downloader::builder(libraries, "output")
        .build()
        .await?;

    // Access YouTube-specific features
    let youtube = downloader.youtube_extractor();
    let results = youtube.search("rust programming", 10).await?;
    let channel = youtube.fetch_channel("UCaYhcUwRBNscFNUKTjgPFiA").await?;

    Ok(())
}
  • 🌐 For any website (YouTube, Vimeo, TikTok, etc.):
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"), 
        PathBuf::from("libs/ffmpeg")
    );
    let downloader = Downloader::builder(libraries, "output")
        .build()
        .await?;

    // Works with any supported site
    let video = downloader.fetch_video_infos("https://vimeo.com/123456789").await?;
    let video_path = downloader.download_video(
        &video,
        "output.mp4"
    ).await?;
    Ok(())
}

Β§πŸ“š Examples

use yt_dlp::Downloader;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let executables_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    // Create fetcher and install binaries
    let downloader = Downloader::with_new_binaries(
        executables_dir,
        output_dir
    ).await?.build().await?;
    
    Ok(())
}
  • πŸ“¦ Installing the yt-dlp binary only:
use yt_dlp::client::deps::LibraryInstaller;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let destination = PathBuf::from("libs");
    let installer = LibraryInstaller::new(destination);

    let youtube = installer.install_youtube(None).await.unwrap();
    Ok(())
}
  • πŸ“¦ Installing the ffmpeg binary only:
use yt_dlp::client::deps::LibraryInstaller;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let destination = PathBuf::from("libs");
    let installer = LibraryInstaller::new(destination);
    
    let ffmpeg = installer.install_ffmpeg(None).await.unwrap();
    Ok(())
}
  • πŸ”„ Updating the yt-dlp binary:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    downloader.update_downloader().await?;
    Ok(())
}
  • πŸ“₯ Fetching a video (with its audio) and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    let video_path = downloader.download_video(&video, "my-video.mp4").await?;
    Ok(())
}
  • πŸ“ Downloading a video to a specific path (ignoring output_dir):
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    
    // Download to an absolute path β€” the file is written directly to the given path,
    // bypassing the configured output_dir.
    let path = PathBuf::from("/Users/me/Videos/my-video.mp4");
    let video_path = downloader.download_video_to_path(&video, path).await?;
    Ok(())
}
  • ✨ Using the fluent API with custom quality preferences:
use yt_dlp::Downloader;
use yt_dlp::model::selector::{VideoQuality, AudioQuality, VideoCodecPreference};
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Use the fluent download builder API
    let video_path = downloader.download(&video, "my-video.mp4")
        .video_quality(VideoQuality::CustomHeight(1080))
        .video_codec(VideoCodecPreference::AVC1)
        .audio_quality(AudioQuality::Best)
        .execute()
        .await?;

    Ok(())
}
  • 🎬 Fetching a video (without its audio) and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    downloader.download_video_stream(&video, "video.mp4").await?;
    Ok(())
}
  • 🎡 Fetching an audio and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    downloader.download_audio_stream(&video, "audio.mp3").await?;
    Ok(())
}
  • πŸ“œ Fetching a specific format and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;
use yt_dlp::VideoSelection;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    println!("Video title: {}", video.title);

    let video_format = video.best_video_format().unwrap();
    let format_path = downloader.download_format(&video_format, "my-video-stream.mp4").await?;
    
    let audio_format = video.worst_audio_format().unwrap();
    let audio_path = downloader.download_format(&audio_format, "my-audio-stream.mp3").await?;
    
    Ok(())
}
  • βš™οΈ Combining an audio and a video file into a single file:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;
use yt_dlp::VideoSelection;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    let audio_format = video.best_audio_format().unwrap();
    let audio_path = downloader.download_format(&audio_format, "audio-stream.mp3").await?;

    let video_format = video.worst_video_format().unwrap();
    let video_path = downloader.download_format(&video_format, "video-stream.mp4").await?;

    let output_path = downloader.combine_audio_and_video("audio-stream.mp3", "video-stream.mp4", "my-output.mp4").await?;
    Ok(())
}
  • πŸ“Έ Fetching a thumbnail and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;
use yt_dlp::model::selector::ThumbnailQuality;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    let thumbnail_path = downloader.download_thumbnail(&video, ThumbnailQuality::Best, "thumbnail.jpg").await?;
    Ok(())
}
  • πŸ–ΌοΈ Selecting a thumbnail by minimum resolution and downloading it:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Best thumbnail by area (width Γ— height)
    if let Some(thumb) = video.best_thumbnail() {
        println!("Best thumbnail: {} β€” {:?}", thumb.url, thumb.resolution);
    }

    // Smallest thumbnail that is at least 1280Γ—720
    if let Some(thumb) = video.thumbnail_for_size(1280, 720) {
        println!("HD thumbnail: {}", thumb.url);
    }

    Ok(())
}
  • πŸ“ Downloading subtitles or automatic captions:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Check available languages (merges subtitles + automatic captions)
    let langs = downloader.list_subtitle_languages(&video);
    println!("Available languages: {:?}", langs);

    // Download French subtitles (falls back to automatic captions if no manual ones)
    let sub_path = downloader.download_subtitle(&video, "fr", "subtitles.srt", true).await?;

    // Download all available subtitles/captions
    let paths = downloader.download_all_subtitles(&video, "subtitles/", true).await?;

    Ok(())
}
  • 🎞️ Downloading storyboard preview frames:
use yt_dlp::Downloader;
use yt_dlp::model::StoryboardQuality;
use yt_dlp::VideoSelection;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Download the best (highest resolution) storyboard into a directory
    let frames = downloader
        .download_storyboard(&video, StoryboardQuality::Best, "storyboard/")
        .await?;
    println!("Downloaded {} MHTML fragment(s)", frames.len());

    // Or pick a specific storyboard format directly
    if let Some(format) = video.best_storyboard_format() {
        let frames = downloader.download_storyboard_format(format, "storyboard/").await?;
    }

    Ok(())
}
  • πŸ“₯ Download with download manager and priority:
use yt_dlp::Downloader;
use yt_dlp::download::manager::{ManagerConfig, DownloadPriority};
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Custom download manager configuration using the typed builder
    let config = ManagerConfig::builder()
        .max_concurrent_downloads(5)        // Maximum 5 concurrent downloads
        .segment_size(1024 * 1024 * 10)    // 10 MB per segment
        .parallel_segments(8)               // 8 parallel segments per download
        .retry_attempts(5)                  // 5 retry attempts on failure
        .max_buffer_size(1024 * 1024 * 20) // 20 MB maximum buffer
        .build();

    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);

    // Create a fetcher with custom configuration
    let downloader = Downloader::with_download_manager_config(libraries, output_dir, config)
        .build()
        .await?;

    // Download a video with high priority
    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    let download_id = downloader.download_video_with_priority(
        &video,
        "video-high-priority.mp4",
        Some(DownloadPriority::High)
    ).await?;

    // Wait for download completion
    let status = downloader.wait_for_download(download_id).await;
    println!("Final download status: {:?}", status);

    Ok(())
}
  • πŸ“Š Download with progress tracking:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Download with progress callback
    let download_id = downloader.download_video_with_progress(
        &video, 
        "video-with-progress.mp4", 
        |downloaded, total| {
            let percentage = if total > 0 {
                (downloaded as f64 / total as f64 * 100.0) as u64
            } else {
                0
            };
            println!("Progress: {}/{} bytes ({}%)", downloaded, total, percentage);
        }
    ).await?;

    // Wait for download completion
    downloader.wait_for_download(download_id).await;
    
    Ok(())
}
  • πŸ›‘ Canceling a download:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Start a download
    let download_id = downloader.download_video_with_priority(
        &video, 
        "video-to-cancel.mp4", 
        None
    ).await?;

    // Check status
    let status = downloader.get_download_status(download_id).await;
    println!("Download status: {:?}", status);

    // Cancel the download
    let canceled = downloader.cancel_download(download_id).await;
    println!("Download canceled: {}", canceled);
    
    Ok(())
}

Β§πŸŽ›οΈ Format Selection

The library provides a powerful format selection system that allows you to download videos and audio with specific quality and codec preferences.

§🎬 Video Quality Options

  • VideoQuality::Best - Selects the highest quality video format available
  • VideoQuality::High - Targets 1080p resolution
  • VideoQuality::Medium - Targets 720p resolution
  • VideoQuality::Low - Targets 480p resolution
  • VideoQuality::Worst - Selects the lowest quality video format available
  • VideoQuality::CustomHeight(u32) - Targets a specific height (e.g., CustomHeight(1440) for 1440p)
  • VideoQuality::CustomWidth(u32) - Targets a specific width (e.g., CustomWidth(1920) for 1920px width)

§🎡 Audio Quality Options

  • AudioQuality::Best - Selects the highest quality audio format available
  • AudioQuality::High - Targets 192kbps bitrate
  • AudioQuality::Medium - Targets 128kbps bitrate
  • AudioQuality::Low - Targets 96kbps bitrate
  • AudioQuality::Worst - Selects the lowest quality audio format available
  • AudioQuality::CustomBitrate(u32) - Targets a specific bitrate in kbps (e.g., CustomBitrate(256) for 256kbps)

§🎞️ Codec Preferences

Β§πŸ“Ή Video Codecs
  • VideoCodecPreference::VP9 - Prefer VP9 codec
  • VideoCodecPreference::AVC1 - Prefer AVC1/H.264 codec
  • VideoCodecPreference::AV1 - Prefer AV01/AV1 codec
  • VideoCodecPreference::Custom(String) - Prefer a custom codec
  • VideoCodecPreference::Any - No codec preference
Β§πŸ”Š Audio Codecs
  • AudioCodecPreference::Opus - Prefer Opus codec
  • AudioCodecPreference::AAC - Prefer AAC codec
  • AudioCodecPreference::MP3 - Prefer MP3 codec
  • AudioCodecPreference::Custom(String) - Prefer a custom codec
  • AudioCodecPreference::Any - No codec preference

Β§πŸ§ͺ Example: Downloading with Quality and Codec Preferences

use yt_dlp::Downloader;
use yt_dlp::model::selector::{VideoQuality, VideoCodecPreference, AudioQuality, AudioCodecPreference};
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");
    
    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");
    
    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Download a high quality video with VP9 codec and high quality audio with Opus codec
    let video_path = downloader.download_video_with_quality(
        &video,
        "complete-video.mp4",
        VideoQuality::High,
        VideoCodecPreference::VP9,
        AudioQuality::High,
        AudioCodecPreference::Opus
    ).await?;
    
    // Download just the video stream with medium quality and AVC1 codec
    let video_stream_path = downloader.download_video_stream_with_quality(
        &video,
        "video-only.mp4",
        VideoQuality::Medium,
        VideoCodecPreference::AVC1
    ).await?;
    
    // Download just the audio stream with high quality and AAC codec
    let audio_stream_path = downloader.download_audio_stream_with_quality(
        &video,
        "audio-only.m4a",
        AudioQuality::High,
        AudioCodecPreference::AAC
    ).await?;
    
    println!("Downloaded files:");
    println!("Complete video: {}", video_path.display());
    println!("Video stream: {}", video_stream_path.display());
    println!("Audio stream: {}", audio_stream_path.display());
    
    Ok(())
}

Β§πŸ“‹ Metadata

The project supports automatic addition of metadata to downloaded files in several formats:

  • MP3: Title, artist, comment, genre (from tags), release year
  • M4A: Title, artist, comment, genre (from tags), release year
  • MP4: All basic metadata, plus technical information (resolution, FPS, video codec, video bitrate, audio codec, audio bitrate, audio channels, sample rate)
  • WebM: All basic metadata (via Matroska format), plus technical information as with MP4
  • FLAC: Title, artist, album, genre, date, description (via Vorbis comments through lofty), thumbnail embedding
  • OGG/Opus: Title, artist, album, genre, date, description (via Vorbis comments through lofty)
  • WAV: Title, artist, album, genre (via RIFF INFO through lofty)
  • AAC: Title, artist, album, genre, date (via ID3v2 through lofty)
  • AIFF: Title, artist, album, genre, date (via ID3v2 through lofty)
  • AVI/TS/FLV: Basic metadata via FFmpeg fallback

Metadata is added automatically during download, without requiring any additional action from the user.

§🧠 Intelligent Metadata Management

The system intelligently manages the application of metadata based on the file type and intended use:

  • For standalone files (audio or audio+video), metadata is applied immediately during download.
  • For separate audio and video streams that will be combined later, metadata is not applied to individual files to avoid redundant work.
  • When combining audio and video streams with combine_audio_and_video(), complete metadata is applied to the final file, including information from both streams.

This optimized approach ensures that metadata is always present in the final file, while avoiding unnecessary processing of temporary files.

Β§πŸ“– Chapters

Videos may contain chapters that divide the content into logical segments. The library provides easy access to chapter information and automatically embeds chapters into downloaded video files (MP4/MKV/WebM):

  • πŸ“– Accessing video chapters:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Check if video has chapters
    if video.has_chapters() {
        println!("Video has {} chapters", video.get_chapters().len());

        // Iterate over all chapters
        for chapter in video.get_chapters() {
            println!(
                "Chapter: {} ({:.2}s - {:.2}s)",
                chapter.title.as_deref().unwrap_or("Untitled"),
                chapter.start_time,
                chapter.end_time
            );
        }
    }

    Ok(())
}
  • πŸ•’ Finding a chapter at a specific timestamp:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Find chapter at 120 seconds (2 minutes)
    if let Some(chapter) = video.get_chapter_at_time(120.0) {
        println!(
            "At 2:00, you're in chapter: {}",
            chapter.title.as_deref().unwrap_or("Untitled")
        );
        println!("Chapter duration: {:.2}s", chapter.duration());
    }

    Ok(())
}

Note: When downloading videos using download_video() or download_video_from_url(), chapters are automatically embedded into the video file metadata. Media players like VLC, MPV, and others will be able to navigate using the chapters!

Β§πŸ”₯ Heatmap

Heatmap data (also known as β€œMost Replayed” segments) shows viewer engagement across different parts of a video. This feature allows you to identify which segments are most popular:

  • πŸ”₯ Accessing heatmap data:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Check if video has heatmap data
    if video.has_heatmap() {
        if let Some(heatmap) = video.get_heatmap() {
            println!("Video has {} heatmap segments", heatmap.points().len());

            // Find the most replayed segment
            if let Some(most_replayed) = heatmap.most_engaged_segment() {
                println!(
                    "Most replayed segment: {:.2}s - {:.2}s (engagement: {:.2})",
                    most_replayed.start_time,
                    most_replayed.end_time,
                    most_replayed.value
                );
            }
        }
    }

    Ok(())
}
  • πŸ“Š Analyzing engagement by threshold:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    if let Some(heatmap) = video.get_heatmap() {
        // Get segments with high engagement (> 0.7)
        let highly_engaged = heatmap.get_highly_engaged_segments(0.7);
        println!("Found {} highly engaged segments", highly_engaged.len());

        for segment in highly_engaged {
            println!(
                "High engagement: {:.2}s - {:.2}s (value: {:.2})",
                segment.start_time,
                segment.end_time,
                segment.value
            );
        }

        // Get engagement at specific timestamp
        if let Some(point) = heatmap.get_point_at_time(120.0) {
            println!(
                "Engagement at 2:00 is {:.2}",
                point.value
            );
        }
    }

    Ok(())
}

Β§πŸ“ Subtitles

The library provides comprehensive subtitle support, including downloading, language selection, and embedding subtitles into videos:

  • πŸ“‹ Listing available subtitle languages:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // List all available subtitle languages
    let languages = downloader.list_subtitle_languages(&video);
    println!("Available subtitle languages: {:?}", languages);

    // Check if specific language is available
    if downloader.has_subtitle_language(&video, "en") {
        println!("English subtitles are available");
    }

    Ok(())
}
  • πŸ“₯ Downloading a specific subtitle:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Download English subtitles
    let subtitle_path = downloader
        .download_subtitle(&video, "en", "subtitle_en.srt", true)
        .await?;
    println!("Subtitle downloaded to: {:?}", subtitle_path);

    Ok(())
}
  • πŸ“₯ Downloading all available subtitles:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir.clone())
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Download all available subtitles
    let subtitle_paths = downloader
        .download_all_subtitles(&video, &output_dir, true)
        .await?;
    println!("Downloaded {} subtitle files", subtitle_paths.len());

    Ok(())
}
  • 🎬 Embedding subtitles into a video:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Download video
    let video_path = downloader.download_video(&video, "video.mp4").await?;

    // Download subtitles
    let en_subtitle = downloader
        .download_subtitle(&video, "en", "subtitle_en.srt", true)
        .await?;
    let fr_subtitle = downloader
        .download_subtitle(&video, "fr", "subtitle_fr.srt", true)
        .await?;

    // Embed subtitles into video
    let video_with_subs = downloader
        .embed_subtitles_in_video(
            &video_path,
            &[en_subtitle, fr_subtitle],
            "video_with_subtitles.mp4",
        )
        .await?;
    println!("Video with embedded subtitles: {:?}", video_with_subs);

    Ok(())
}
  • πŸ”„ Working with automatic captions:
use yt_dlp::Downloader;
use yt_dlp::model::caption::Subtitle;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // Iterate over subtitles and filter automatic ones
    for (lang_code, subtitles) in &video.subtitles {
        for subtitle in subtitles {
            if subtitle.is_automatic {
                println!(
                    "Auto-generated subtitle: {} ({})",
                    subtitle.language_name
                        .as_deref()
                        .unwrap_or(lang_code),
                    subtitle.file_extension()
                );
            }
        }
    }

    // Convert automatic captions to Subtitle struct
    for (lang_code, auto_captions) in &video.automatic_captions {
        if let Some(caption) = auto_captions.first() {
            let subtitle = Subtitle::from_automatic_caption(
                caption,
                lang_code.clone(),
            );
            println!("Converted: {}", subtitle);
        }
    }

    Ok(())
}

Β§πŸ“‚ Playlists

The library provides full playlist support, including fetching playlist metadata and downloading videos with various selection options:

  • πŸ“‹ Fetching playlist information:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let playlist_url = String::from("https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf");
    let playlist = downloader.fetch_playlist_infos(playlist_url).await?;

    println!("Playlist: {}", playlist.title);
    println!("Videos: {}", playlist.entry_count());
    println!("Uploader: {}", playlist.uploader.as_deref().unwrap_or("unknown"));

    // List all videos in the playlist
    for entry in &playlist.entries {
        println!(
            "[{}] {} ({})",
            entry.index.unwrap_or(0),
            entry.title,
            entry.id
        );
    }

    Ok(())
}
  • πŸ“₯ Downloading entire playlist:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let playlist_url = String::from("https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf");
    let playlist = downloader.fetch_playlist_infos(playlist_url).await?;

    // Download all videos with a pattern
    // Use %(playlist_index)s for index, %(title)s for title, %(id)s for video ID
    let video_paths = downloader
        .download_playlist(&playlist, "%(playlist_index)s - %(title)s.mp4")
        .await?;

    println!("Downloaded {} videos", video_paths.len());

    Ok(())
}
  • 🎯 Downloading specific videos by index:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let playlist_url = String::from("https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf");
    let playlist = downloader.fetch_playlist_infos(playlist_url).await?;

    // Download specific videos by index (0-based)
    let indices = vec![0, 2, 5, 10]; // Videos at positions 1, 3, 6, and 11
    let video_paths = downloader
        .download_playlist_items(&playlist, &indices, "%(playlist_index)s - %(title)s.mp4")
        .await?;

    println!("Downloaded {} specific videos", video_paths.len());

    Ok(())
}
  • πŸ“Š Downloading a range of videos:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let playlist_url = String::from("https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf");
    let playlist = downloader.fetch_playlist_infos(playlist_url).await?;

    // Download videos 5-15 (0-based, inclusive)
    let video_paths = downloader
        .download_playlist_range(&playlist, 5, 15, "%(playlist_index)s - %(title)s.mp4")
        .await?;

    println!("Downloaded {} videos from range", video_paths.len());

    Ok(())
}
  • πŸ” Filtering and analyzing playlists:
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let youtube = libraries_dir.join("yt-dlp");
    let ffmpeg = libraries_dir.join("ffmpeg");

    let libraries = Libraries::new(youtube, ffmpeg);
    let downloader = Downloader::builder(libraries, output_dir)
        .build()
        .await?;

    let playlist_url = String::from("https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf");
    let playlist = downloader.fetch_playlist_infos(playlist_url).await?;

    // Check if playlist is complete
    if playlist.is_complete() {
        println!("All playlist videos have been fetched");
    }

    // Get only available videos
    let available = playlist.available_entries();
    println!("Available videos: {}/{}", available.len(), playlist.entry_count());

    // Get specific entry
    if let Some(first_video) = playlist.get_entry_by_index(0) {
        println!("First video: {}", first_video.title);
        if let Some(duration) = first_video.duration_minutes() {
            println!("Duration: {:.2} minutes", duration);
        }
    }

    // Get entries in a range
    let range = playlist.get_entries_in_range(0, 10);
    println!("First 11 videos: {}", range.len());

    Ok(())
}

Β§πŸ”” Events, Hooks & Webhooks

The library provides a comprehensive event system to monitor download lifecycle and react to events through Rust hooks or HTTP webhooks.

§⚑ Event System

All download operations emit events that you can subscribe to:

  • πŸ“‘ Subscribing to the event stream:
use yt_dlp::Downloader;
use tokio_stream::StreamExt;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    let downloader = Downloader::builder(libraries, output_dir).build().await?;
    let mut stream = downloader.event_stream();

    while let Some(Ok(event)) = stream.next().await {
        println!("Event: {} - {:?}", event.event_type(), event);
    }

    Ok(())
}

Β§βœ‰οΈ Available Events

The library emits 22 different event types covering the entire download lifecycle:

Download Lifecycle:

  • VideoFetched - Video metadata retrieved
  • DownloadQueued - Download added to queue
  • DownloadStarted - Download begins
  • DownloadProgress - Progress updates (bytes downloaded, speed, ETA)
  • DownloadPaused / DownloadResumed - Pause/resume events
  • DownloadCompleted - Download finished successfully
  • DownloadFailed - Download failed with error
  • DownloadCanceled - Download was canceled

Format & Metadata:

  • FormatSelected - Video/audio format chosen
  • MetadataApplied - Metadata tags written
  • ChaptersEmbedded - Chapters added to file

Post-Processing:

  • PostProcessStarted / PostProcessCompleted / PostProcessFailed - FFmpeg operations

Playlist Operations:

  • PlaylistFetched - Playlist metadata retrieved
  • PlaylistItemStarted / PlaylistItemCompleted / PlaylistItemFailed - Per-item events
  • PlaylistCompleted - Entire playlist finished

Advanced:

  • SegmentStarted / SegmentCompleted - Parallel segment downloads

Β§πŸͺ Rust Hooks (Feature: hooks)

Register async functions to be called when events occur:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["hooks"] }
  • 🎣 Registering a hook for download events:
β“˜
use yt_dlp::events::{EventHook, EventFilter, DownloadEvent, HookResult};
use async_trait::async_trait;
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[derive(Clone)]
struct MyHook;

#[async_trait]
impl EventHook for MyHook {
    async fn on_event(&self, event: &DownloadEvent) -> HookResult {
        match event {
            DownloadEvent::DownloadCompleted { download_id, output_path, .. } => {
                println!("Download {} completed: {:?}", download_id, output_path);
            }
            DownloadEvent::DownloadFailed { download_id, error, .. } => {
                eprintln!("Download {} failed: {}", download_id, error);
            }
            _ => {}
        }
        Ok(())
    }

    fn filter(&self) -> EventFilter {
        // Only receive terminal events (completed, failed, canceled)
        EventFilter::only_terminal()
    }
}

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    let mut downloader = Downloader::builder(libraries, output_dir).build().await?;
    downloader.register_hook(MyHook).await;

    Ok(())
}

Hook Features:

  • Async execution
  • Event filtering (by type, download ID, custom predicates)
  • Parallel or sequential execution
  • Automatic timeout protection (30s)
  • Error isolation (hook failures don’t stop downloads)
Β§πŸ” Event Filters
use yt_dlp::events::EventFilter;

// Only completed downloads
EventFilter::only_completed();

// Only failed downloads
EventFilter::only_failed();

// Progress updates only
EventFilter::only_progress();

// Exclude progress events
EventFilter::all().exclude_progress();

// Specific download ID
EventFilter::download_id(123);

// Terminal events (completed, failed, canceled)
EventFilter::only_terminal();

// Chain filters
EventFilter::download_id(123).and_then(|e| e.is_terminal());

// Custom filter
EventFilter::all().and_then(|event| {
    // Your custom logic
    true
});

Β§πŸ“‘ HTTP Webhooks (Feature: webhooks)

Send events to external HTTP endpoints with automatic retry:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["webhooks"] }
  • πŸ“‘ Registering a webhook:
β“˜
use yt_dlp::events::{WebhookConfig, WebhookMethod, EventFilter};
use std::time::Duration;
use yt_dlp::Downloader;
use std::path::PathBuf;
use yt_dlp::client::deps::Libraries;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    let webhook = WebhookConfig::new("https://example.com/webhook")
        .with_method(WebhookMethod::Post)
        .with_header("Authorization", "Bearer your-token")
        .with_filter(EventFilter::only_completed())
        .with_timeout(Duration::from_secs(10));

    let mut downloader = Downloader::builder(libraries, output_dir).build().await?;
    downloader.register_webhook(webhook).await;

    Ok(())
}

Webhook Features:

  • HTTP POST/PUT/PATCH methods
  • Custom headers (authentication, etc.)
  • Event filtering (same as hooks)
  • Automatic retry with exponential backoff (3 attempts by default)
  • Configurable timeouts
  • JSON payload with event data
  • Environment variable configuration

Environment Variables:

export YTDLP_WEBHOOK_URL="https://example.com/webhook"
export YTDLP_WEBHOOK_METHOD="POST"  # Optional, default: POST
export YTDLP_WEBHOOK_TIMEOUT="10"   # Optional, default: 10 seconds
  • πŸ”§ Loading a webhook from environment variables:
β“˜
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use yt_dlp::events::WebhookConfig;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg")
    );
    let mut downloader = Downloader::builder(libraries, PathBuf::from("output")).build().await?;

    // Load webhook from environment variables
    if let Some(webhook) = WebhookConfig::from_env() {
        downloader.register_webhook(webhook).await;
    }

    Ok(())
}

Webhook Payload:

{
  "event_type": "download_completed",
  "download_id": 123,
  "timestamp": "2025-01-21T10:30:00Z",
  "data": {
    "download_id": 123,
    "output_path": "/path/to/video.mp4",
    "duration": 45.2,
    "total_bytes": 104857600
  }
}
§♻️ Retry Strategy
  • ♻️ Configuring a retry strategy:
β“˜
use yt_dlp::events::RetryStrategy;
use std::time::Duration;

// Exponential backoff (default)
let strategy = RetryStrategy::exponential(
    3,                              // max attempts
    Duration::from_secs(1),         // initial delay
    Duration::from_secs(30)         // max delay
);

// Linear backoff
let strategy = RetryStrategy::linear(
    3,                              // max attempts
    Duration::from_secs(5)          // fixed delay
);

// No retries
let strategy = RetryStrategy::none();

Β§πŸ”— Combining Hooks and Webhooks

Use both hooks and webhooks together:

  • πŸ”— Using hooks and webhooks simultaneously:
β“˜
use yt_dlp::Downloader;
use yt_dlp::events::{EventHook, WebhookConfig, EventFilter};
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[derive(Clone)]
struct MyLocalHook;

#[async_trait::async_trait]
impl EventHook for MyLocalHook {
    async fn on_event(&self, _event: &yt_dlp::events::DownloadEvent) -> yt_dlp::events::HookResult { Ok(()) }
    fn filter(&self) -> EventFilter { EventFilter::all() }
}

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(PathBuf::from("libs/yt-dlp"), PathBuf::from("libs/ffmpeg"));

    let mut downloader = Downloader::builder(libraries, PathBuf::from("output")).build().await?;

    // Register Rust hook for immediate in-process handling
    downloader.register_hook(MyLocalHook).await;

    // Register webhook for external notifications
    let webhook = WebhookConfig::new("https://example.com/webhook")
        .with_filter(EventFilter::only_completed());
    downloader.register_webhook(webhook).await;

    // Start downloads - both hooks and webhooks will receive events
    let video = downloader.fetch_video_infos("https://youtube.com/watch?v=...".to_string()).await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}

Β§πŸ“Š Statistics & Analytics (Feature: statistics)

Enable real-time, aggregate metrics with zero manual bookkeeping:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["statistics"] }

The StatisticsTracker subscribes to the internal event bus in a background task and continuously updates running counters. Call snapshot() at any time to obtain an atomic view of all metrics:

β“˜
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(PathBuf::from("libs/yt-dlp"), PathBuf::from("libs/ffmpeg"));
    let downloader = Downloader::builder(libraries, "output").build().await?;

    // Perform some downloads and fetches ...
    let video = downloader.fetch_video_infos("https://youtube.com/watch?v=...".to_string()).await?;
    downloader.download_video(&video, "video.mp4").await?;

    let snapshot = downloader.statistics().snapshot().await;
    println!("Downloads completed:  {}", snapshot.downloads.completed);
    println!("Total bytes:          {}", snapshot.downloads.total_bytes);
    println!("Avg speed (B/s):      {:?}", snapshot.downloads.avg_speed_bytes_per_sec);
    println!("Download success %:   {:?}", snapshot.downloads.success_rate);
    println!("Fetch success %:      {:?}", snapshot.fetches.success_rate);
    println!("Post-process success: {:?}", snapshot.post_processing.success_rate);

    Ok(())
}

The snapshot exposes:

  • downloads β€” attempted, completed, failed, canceled, total bytes, average speed, peak speed, success rate
  • fetches β€” attempted, succeeded, failed, average duration, success rate (video + playlist fetches)
  • post_processing β€” attempted, succeeded, failed, average duration
  • playlists β€” playlists fetched, failed, per-item success rate
  • recent_downloads β€” bounded history window of completed downloads with per-download details

Β§πŸš€ Advanced Features

Β§πŸ” Proxy Support

The library supports HTTP, HTTPS, and SOCKS5 proxies for both yt-dlp and reqwest downloads:

  • πŸ” Configuring a proxy with authentication:
use yt_dlp::Downloader;
use yt_dlp::client::proxy::{ProxyConfig, ProxyType};
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    // Configure proxy with authentication
    let proxy = ProxyConfig::new(ProxyType::Http, "http://proxy.example.com:8080")
        .with_auth("username", "password")
        .with_no_proxy(vec!["localhost".to_string(), "127.0.0.1".to_string()]);

    // Build Downloader with proxy
    let downloader = Downloader::builder(libraries, output_dir)
        .with_proxy(proxy)
        .build()
        .await?;

    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;

    // All downloads (video, audio, thumbnails) will use the proxy
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}

Supported proxy types:

  • HTTP/HTTPS: Standard HTTP proxies
  • SOCKS5: SOCKS5 proxies for more flexibility
  • Authentication: Username/password authentication
  • No-proxy list: Exclude specific domains from proxying

Β§πŸ”‘ Authentication & Cookies

Many platforms (YouTube bot-protection, Twitch, age-restricted content, etc.) require authentication. The library supports three authentication modes that are propagated to both metadata extraction and all download operations automatically.

use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );

    // Export cookies from your browser with a browser extension (e.g. "Get cookies.txt LOCALLY")
    let downloader = Downloader::builder(libraries, PathBuf::from("output"))
        .with_cookies("cookies.txt")
        .build()
        .await?;

    let video = downloader.fetch_video_infos("https://www.youtube.com/watch?v=gXtp6C-3JKo").await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}
Β§Browser cookies
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );

    // yt-dlp will read cookies directly from your browser's cookie store
    let downloader = Downloader::builder(libraries, PathBuf::from("output"))
        .with_cookies_from_browser("chrome") // or "firefox", "safari", "edge", …
        .build()
        .await?;

    let video = downloader.fetch_video_infos("https://www.youtube.com/watch?v=gXtp6C-3JKo").await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}
Β§Runtime (after build)
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );

    let mut downloader = Downloader::builder(libraries, PathBuf::from("output"))
        .build()
        .await?;

    // Apply cookies after build β€” propagates to both extractors and download args
    downloader.set_cookies("cookies.txt");
    // or: downloader.set_cookies_from_browser("chrome");
    // or: downloader.set_netrc();

    let video = downloader.fetch_video_infos("https://www.youtube.com/watch?v=gXtp6C-3JKo").await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}
Β§.netrc
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );

    let downloader = Downloader::builder(libraries, PathBuf::from("output"))
        .with_netrc()
        .build()
        .await?;

    let video = downloader.fetch_video_infos("https://www.youtube.com/watch?v=gXtp6C-3JKo").await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}

Β§βœ‚οΈ Clip extraction & chapter splitting

The library supports downloading a specific time range or a specific chapter from a video without fetching the whole file. Seeking is handled by media-seek β€” a pure Rust container index parser that translates timestamps into HTTP Range byte offsets.

  • βœ‚οΈ Downloading a specific time range (seconds):
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Download seconds [60, 120] only β€” no re-encoding
    let clip_path = downloader
        .download(&video, "clip.mp4")
        .time_range(60.0, 120.0)?
        .execute()
        .await?;

    Ok(())
}
  • πŸ“– Downloading a specific chapter by index range:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Download chapters 0 through 2 (inclusive)
    let clip_path = downloader
        .download(&video, "chapters.mp4")
        .chapters(0, 2)?
        .execute()
        .await?;

    Ok(())
}
  • πŸ”ͺ Splitting a downloaded video into one file per chapter:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let url = "https://www.youtube.com/watch?v=gXtp6C-3JKo";
    let video = downloader.fetch_video_infos(url).await?;

    // Download and split into one file per chapter β€” FFmpeg stream copy, no re-encoding
    let chapter_files: Vec<PathBuf> = downloader
        .split_by_chapters(&video, "output/chapters/")
        .await?;

    for path in &chapter_files {
        println!("Chapter file: {}", path.display());
    }

    Ok(())
}

Β§πŸ”΄ Live Stream Recording (Feature: live-recording)

Record live streams using either the pure-Rust reqwest engine or FFmpeg as a fallback. Enable the feature in your Cargo.toml:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["live-recording"] }
Β§πŸ“₯ Basic live recording (reqwest engine)
β“˜
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;
use std::time::Duration;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let video = downloader.fetch_video_infos("https://youtube.com/watch?v=LIVE_ID").await?;

    // Record for 1 hour maximum
    let result = downloader.record_live(&video, "live-recording.ts")
        .with_max_duration(Duration::from_secs(3600))
        .execute()
        .await?;

    println!("Recorded {} bytes in {:.1}s", result.total_bytes, result.total_duration.as_secs_f64());
    Ok(())
}
§🎬 FFmpeg fallback engine
β“˜
use yt_dlp::Downloader;
use yt_dlp::events::RecordingMethod;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;
use std::time::Duration;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let video = downloader.fetch_video_infos("https://youtube.com/watch?v=LIVE_ID").await?;

    let result = downloader.record_live(&video, "live-recording.ts")
        .with_method(RecordingMethod::Fallback)
        .with_max_duration(Duration::from_secs(600))
        .execute()
        .await?;

    Ok(())
}

Implementation details:

  • Reqwest engine (default): Pure-Rust HLS segment fetcher. Polls the media playlist, downloads new segments, and writes them sequentially. Zero-copy bytes::Bytes, progress events throttled at 50 ms.
  • FFmpeg engine (fallback): Spawns ffmpeg -i <url> -c copy <output>. Stops gracefully via stdin q. Useful for encrypted streams or complex HLS features.
  • Recording stops on cancellation token, #EXT-X-ENDLIST, or max duration.
  • Recording events: LiveRecordingStarted, LiveRecordingProgress, LiveRecordingStopped, LiveRecordingFailed.
  • Streaming events: LiveStreamStarted, LiveStreamProgress, LiveStreamStopped, LiveStreamFailed.
Β§πŸ“‘ Live fragment streaming (Feature: live-streaming)

Enable the feature in your Cargo.toml:

[dependencies]
yt-dlp = { version = "2.7.0", features = ["live-streaming"] }
β“˜
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;
use tokio_stream::StreamExt;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg"),
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;

    let video = downloader.fetch_video_infos("https://youtube.com/watch?v=LIVE_ID").await?;

    let mut stream = downloader.stream_live(&video)
        .execute()
        .await?;

    while let Some(fragment) = stream.next().await {
        let fragment = fragment?;
        println!("Fragment {} bytes", fragment.data.len());
    }

    Ok(())
}

§🎨 Post-Processing Options

Apply advanced post-processing to videos using FFmpeg:

Β§πŸ”§ Basic codec conversion
  • πŸ”§ Converting video codec and bitrate:
use yt_dlp::Downloader;
use yt_dlp::download::{PostProcessConfig, VideoCodec, AudioCodec};
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );
    let downloader = Downloader::builder(libraries, output_dir).build().await?;

    // Configure post-processing
    let config = PostProcessConfig::new()
        .with_video_codec(VideoCodec::H264)
        .with_audio_codec(AudioCodec::AAC)
        .with_video_bitrate("2M")
        .with_audio_bitrate("192k");

    // Apply to existing video
    downloader.postprocess_video("input.mp4", "output.mp4", config).await?;

    Ok(())
}
Β§πŸŽ›οΈ Advanced post-processing with filters
  • πŸŽ›οΈ Applying resolution, framerate, and visual filters:
use yt_dlp::Downloader;
use yt_dlp::download::{
    PostProcessConfig, VideoCodec, Resolution, EncodingPreset,
    FfmpegFilter, WatermarkPosition
};
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );
    let downloader = Downloader::builder(libraries, output_dir).build().await?;

    // Advanced configuration with filters
    let config = PostProcessConfig::new()
        .with_video_codec(VideoCodec::H265)
        .with_resolution(Resolution::HD)
        .with_framerate(30)
        .with_preset(EncodingPreset::Medium)
        .add_filter(FfmpegFilter::Brightness { value: 0.1 })
        .add_filter(FfmpegFilter::Contrast { value: 1.2 })
        .add_filter(FfmpegFilter::Watermark {
            path: "logo.png".to_string(),
            position: WatermarkPosition::BottomRight,
        });

    downloader.postprocess_video("input.mp4", "processed.mp4", config).await?;

    Ok(())
}
Β§πŸ“‹ Available post-processing options

Video Codecs:

  • H.264 (libx264) - Most compatible
  • H.265 (libx265) - Better compression
  • VP9 (libvpx-vp9) - Open format
  • AV1 (libaom-av1) - Next-gen codec
  • Copy - No re-encoding

Audio Codecs:

  • AAC - High quality, widely supported
  • MP3 (libmp3lame) - Universal compatibility
  • Opus - Best quality/size ratio
  • Vorbis - Open format
  • Copy - No re-encoding

Resolutions:

  • UHD8K (7680x4320)
  • UHD4K (3840x2160)
  • QHD (2560x1440)
  • FullHD (1920x1080)
  • HD (1280x720)
  • SD (854x480)
  • Low (640x360)
  • Custom { width, height }

Encoding Presets:

  • UltraFast, SuperFast, VeryFast, Fast
  • Medium (balanced)
  • Slow, Slower, VerySlow (best quality)

Video Filters:

  • Crop: Crop { width, height, x, y }
  • Rotate: Rotate { angle } (in degrees)
  • Watermark: Watermark { path, position }
  • Brightness: Brightness { value } (-1.0 to 1.0)
  • Contrast: Contrast { value } (0.0 to 4.0)
  • Saturation: Saturation { value } (0.0 to 3.0)
  • Blur: Blur { radius }
  • FlipHorizontal, FlipVertical
  • Denoise, Sharpen
  • Custom: Custom { filter } - Any FFmpeg filter string

§⚑ Speed Profiles

The library includes an intelligent speed optimization system that automatically configures download parameters based on your internet connection speed. This feature significantly improves download performance for both individual videos and playlists.

Β§πŸ“Š Available Speed Profiles

Three pre-configured profiles are available:

🐒 Conservative (for connections < 50 Mbps)

  • 3 concurrent downloads
  • 4-8 parallel segments per file
  • 5 MB segment size
  • 10 MB buffer
  • 2 concurrent playlist downloads
  • Best for: Standard internet, avoiding network congestion, limited bandwidth

βš–οΈ Balanced (for connections 50-500 Mbps) - Default

  • 4 concurrent downloads
  • 5–20 parallel segments per file
  • 8 MB segment size
  • 20 MB buffer
  • 3 concurrent playlist downloads
  • Best for: Most modern internet connections, general use

πŸš€ Aggressive (for connections > 500 Mbps)

  • 6 concurrent downloads
  • 6–24 parallel segments per file
  • 10 MB segment size
  • 30 MB buffer
  • 5 concurrent playlist downloads
  • Best for: High-bandwidth connections (fiber, gigabit), maximum speed
Β§πŸš€ Using Speed Profiles
  • πŸš€ Selecting a speed profile at build time:
use yt_dlp::Downloader;
use yt_dlp::download::SpeedProfile;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    // Use the Aggressive profile for maximum speed
    let downloader = Downloader::builder(libraries, output_dir)
        .with_speed_profile(SpeedProfile::Aggressive)
        .build()
        .await?;

    // All downloads will now use optimized settings
    let url = String::from("https://www.youtube.com/watch?v=gXtp6C-3JKo");
    let video = downloader.fetch_video_infos(url).await?;
    downloader.download_video(&video, "video.mp4").await?;

    Ok(())
}
Β§βš™οΈ Manual Configuration
  • βš™οΈ Fine-grained control over download parameters:
use yt_dlp::Downloader;
use yt_dlp::download::ManagerConfig;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
pub async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries_dir = PathBuf::from("libs");
    let output_dir = PathBuf::from("output");

    let libraries = Libraries::new(
        libraries_dir.join("yt-dlp"),
        libraries_dir.join("ffmpeg")
    );

    // Create a custom configuration
    let config = ManagerConfig::builder()
        .max_concurrent_downloads(10)   // 10 concurrent downloads
        .segment_size(15 * 1024 * 1024) // 15 MB segments
        .parallel_segments(16)          // 16 parallel segments
        .build();

    let downloader = Downloader::builder(libraries, output_dir)
        .with_download_manager_config(config)
        .build()
        .await?;

    Ok(())
}
Β§πŸ“ˆ Performance Improvements

The speed optimization system includes several advanced features:

  • HTTP/2 Support: Automatically enabled for better connection multiplexing
  • Transparent Compression: Responses are automatically decompressed (gzip, brotli) for faster transfers
  • TCP Nodelay: Nagle’s algorithm is disabled for lower latency on small writes
  • CDN-Friendly Range Probing: Uses GET with Range: bytes=0-0 instead of HEAD for better CDN compatibility
  • Progress Throttling: Progress callbacks are throttled to 50 ms intervals to reduce overhead
  • Parallel Playlist Downloads: Playlists are downloaded in parallel by default (previously sequential)
  • Dynamic Segment Allocation: Automatically adjusts the number of parallel segments based on file size
  • Connection Pooling: Reuses HTTP connections for better performance
  • Intelligent Buffering: Optimized buffer sizes based on your profile
  • Async DNS (opt-in): Enable the hickory-dns feature for a fully async, pure-Rust DNS resolver

Expected Performance Gains:

For individual videos:

  • Conservative: ~30% faster (HTTP/2)
  • Balanced: ~100% faster (2x segments + HTTP/2)
  • Aggressive: ~200% faster (3x segments + HTTP/2)

For playlists:

  • Conservative: ~150% faster (2 videos in parallel)
  • Balanced: ~200% faster (3 videos in parallel)
  • Aggressive: ~400% faster (5 videos in parallel)

Note: Actual performance gains depend on your internet speed, server limitations, and network conditions.

§🌍 Multi-Site Support

This library supports all 1,800+ extractors from yt-dlp!

While the examples focus on YouTube (the most common use case), the library works seamlessly with any site supported by yt-dlp. Simply pass the URL - yt-dlp automatically detects the correct extractor.

Β§πŸ—‚οΈ Supported Sites

  • Video platforms: YouTube, Vimeo, Dailymotion, Twitch
  • Social media: Instagram, TikTok, Twitter/X, Facebook
  • Streaming services: Netflix, Disney+, Crunchyroll (may require authentication)
  • Music platforms: Spotify, SoundCloud, Bandcamp
  • News outlets: CNN, BBC, Fox News
  • And 1,800+ more…

For the complete list, see yt-dlp’s supported sites.

§🧩 Examples

  • 🎬 Downloading from Vimeo:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(PathBuf::from("libs/yt-dlp"), PathBuf::from("libs/ffmpeg"));
    let downloader = Downloader::builder(libraries, PathBuf::from("output")).build().await?;

    let url = "https://vimeo.com/148751763".to_string();
    let video = downloader.fetch_video_infos(url).await?;
    downloader.download_video(&video, "vimeo-video.mp4").await?;

    Ok(())
}
  • πŸ“± Downloading from TikTok:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(PathBuf::from("libs/yt-dlp"), PathBuf::from("libs/ffmpeg"));
    let downloader = Downloader::builder(libraries, PathBuf::from("output")).build().await?;

    let url = "https://www.tiktok.com/@user/video/123".to_string();
    let video = downloader.fetch_video_infos(url).await?;
    downloader.download_video(&video, "tiktok-video.mp4").await?;

    Ok(())
}
  • πŸ“Έ Downloading from Instagram (may require cookies for authentication):
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(PathBuf::from("libs/yt-dlp"), PathBuf::from("libs/ffmpeg"));
    let downloader = Downloader::builder(libraries, PathBuf::from("output")).build().await?;

    let url = "https://www.instagram.com/p/ABC123/".to_string();
    let video = downloader.fetch_video_infos(url).await?;

    Ok(())
}
  • πŸ” Detecting which extractor handles a URL:
use yt_dlp::Downloader;
use yt_dlp::client::deps::Libraries;
use std::path::PathBuf;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let libraries = Libraries::new(
        PathBuf::from("libs/yt-dlp"),
        PathBuf::from("libs/ffmpeg")
    );
    let downloader = Downloader::builder(libraries, "output").build().await?;
    let extractor = downloader.detect_extractor("https://vimeo.com/123").await?;
    println!("This URL uses the '{}' extractor", extractor);
    Ok(())
}

For detailed documentation, examples, and authentication instructions, see the extractor module documentation.

Β§πŸ”¬ Profiling (Feature: profiling)

See PROFILING.md for the complete guide (flamegraph, samply, dhat-rs, heaptrack, Criterion, and the raw vs library comparison tool).


§🏎️ Performances

The library fetches video metadata via yt-dlp --dump-single-json, then downloads format streams directly over HTTP using parallel segments β€” bypassing yt-dlp’s sequential download engine. Run benches/compare.rs with any public YouTube URL to reproduce these numbers on your own connection:

cargo bench --bench compare --features profiling -- https://www.youtube.com/watch?v=gXtp6C-3JKo --cookies-from-browser safari --runs 10

Results below are averages over 10 runs on a typical broadband connection.

Methodology: raw yt-dlp re-fetches metadata on every download. The library fetches metadata once, caches it, and then downloads each format via parallel HTTP segments β€” this separation is the core optimisation reflected in the results below.

§🎡 Audio streams

Scenarioyt-dlpConservativeBalanced (default)Aggressive
Audio 96 kbps (Low)8.26s1.27s1.47s1.27s
Audio 128 kbps (Medium)7.83s1.11s1.12s1.18s
Audio 192 kbps (High)8.53s1.38s1.26s1.19s
Audio best quality8.51s1.22s1.15s1.15s

§🎬 Video streams (no audio)

Scenarioyt-dlpConservativeBalanced (default)Aggressive
Video 480p8.59s2.39s2.11s2.58s
Video 720p9.84s3.67s3.86s3.89s
Video 1080p17.6s8.65s8.60s8.81s
Video best quality16.6s11.7s11.8s12.0s

Β§πŸ“¦ Muxed streams β€” native (YouTube pre-muxed, no ffmpeg)

YouTube serves some formats already containing both video and audio tracks. No post-processing is needed β€” the file is downloaded as-is.

Scenarioyt-dlpConservativeBalanced (default)Aggressive
Native 360p (mp4)9.69s2.11s2.00s2.16s
Native 720p (mp4)20.7s2.08s2.10s2.06s

Β§πŸ“¦ Muxed streams β€” combined by ffmpeg

For higher-quality streams, YouTube only provides separate video and audio tracks. The library downloads both in parallel, then ffmpeg merges them using stream copy (no re-encoding) when the audio container is compatible with the output format (e.g. AAC/m4a β†’ mp4).

Scenarioyt-dlpConservativeBalanced (default)Aggressive
Muxed 480p9.41s3.51s3.48s3.34s
Muxed 720p10.4s4.95s4.92s4.97s
Muxed 1080p18.3s10.4s10.5s10.3s
Muxed best quality18.2s13.7s13.7s13.3s

Β§πŸš€ Speed profiles

ProfileParallel segmentsSegment sizeUse case
Conservative1–165 MB< 50 Mbps connections
Balanced (default)2–208 MBMost modern connections
Aggressive3–2410 MBFibre / gigabit

See PROFILING.md for detailed micro-benchmarks.


Β§πŸ’‘Features coming soon

  • Use rust-ffmpeg and ffmpeg-sys-next as safe bindings instead of commands, and keep Command fallback as a feature flag
  • Bandwidth throttling (limit download speed)
  • Download queue persistence (resume queue across restarts)
  • SponsorBlock integration (skip/mark sponsor segments)

Β§πŸ” The media-seek crate

This library uses media-seek, a standalone sub-crate published independently on crates.io, for container index parsing. When you use clip extraction or chapter range downloads, media-seek parses the stream header to find the exact byte offsets β€” no subprocess, no FFmpeg involved for the seek step.

Supported container formats: MP4/M4A (fMP4 SIDX), WebM/MKV (EBML Cues), MP3 (Xing/VBRI TOC or CBR), OGG (granule bisection), FLAC (SEEKTABLE), WAV, AIFF, AAC/ADTS, FLV (AMF0 keyframes), AVI (idx1), MPEG-TS (PCR binary search).

You can also use media-seek directly in your own project:

[dependencies]
media-seek = "0.4.0"

See the media-seek README for the full API reference and usage examples.


§🀝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. Make sure to follow the Contributing Guidelines.

Β§πŸ“„ License

This project is licensed under the GPL-3.0 License.

Re-exportsΒ§

pub use metadata::PlaylistMetadata;
pub use client::streams::selection::VideoSelection;
pub use client::DownloadBuilder;
pub use client::DownloaderBuilder;
pub use download::DownloadManager;
pub use download::DownloadPriority;
pub use download::DownloadStatus;
pub use model::utils::AllTraits;
pub use model::utils::CommonTraits;

ModulesΒ§

cache
Cache module for storing video metadata and downloaded files.
client
Downloader client module.
download
Download orchestration module.
error
Error types with enhanced context and structured information.
events
Event system for download lifecycle notifications
executor
Command execution module.
extractor
Video extractor system for multi-site support.
live
Live stream recording and streaming module.
macros
Convenience macros for common operations.
metadata
Metadata management module for downloaded files.
model
The models used to represent the data fetched by β€˜yt-dlp’.
prelude
Prelude module for convenient imports.
stats
Statistics and analytics for download and metadata fetch operations.
utils
Utility functions and types used throughout the application.

MacrosΒ§

install_libraries
Create a Libraries instance with automatic binary installation.
simple_hook
Helper macro for creating simple hooks from closures
ternary
A macro to mimic the ternary operator in Rust.
youtube
Create a Youtube instance with sensible defaults.
ytdlp_args
Configure yt-dlp arguments easily.

StructsΒ§

Downloader
Universal video downloader supporting 1,800+ sites via yt-dlp.