Expand description
π¬οΈ A Rust library (with auto dependencies installation) for video downloading
Β§ποΈ 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However, I quickly realized that due to frequent breaking changes on the YouTube website, the crate was outdated and no longer functional.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§π₯ 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§π 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.
- πͺ
hooks- Enables Rust hooks and callbacks for download events. Allows registering async functions that will be called when events occur. - π‘
webhooks- Enables HTTP webhooks delivery for download events. Allows sending events to external HTTP endpoints with retry logic. - π
statistics- Enables real-time statistics and analytics on downloads and fetches. Exposes aggregate counters, averages, success rates, and a bounded history window. - β‘
cache-memory(enabled by default) β In-memory Moka cache (pulls inmoka). Fast TTL-based eviction; no persistence. - ποΈ
cache-jsonβ JSON file-system backend. One.jsonfile per entry. - ποΈ
cache-redbβ Embeddedredbbackend. Single-file, pure-rust, ACID-compliant. - π
cache-redisβ DistributedRedisbackend. Native TTL viaSETEX. - π΄
live-recording- Enables live stream recording via HLS segment fetching (reqwest) or FFmpeg fallback. Pulls inm3u8-rsfor HLS manifest parsing. - π‘
live-streaming- Enables live fragment streaming via HLS segment fetching (reqwest). Pulls inm3u8-rsfor HLS manifest parsing. - π
rustls- Enables therustls-tlsfeature in thereqwestcrate. This enables building the application without openssl or other system sourced SSL libraries. - π
hickory-dns- Enables async DNS resolution viaHickory DNS(passesreqwest/hickory-dns). Replaces the default blocking system resolver with a fully async, pure-Rust resolver.
Β§ποΈ 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:
| Feature | Backend | Persistence | Notes |
|---|---|---|---|
cache-memory (default) | In-memory Moka | β No | TTL-based eviction, async-ready |
cache-json | JSON files on disk | β Yes | One .json file per entry in the cache directory |
cache-redb | Embedded redb | β Yes | Single-file, pure-Rust, ACID transactions |
cache-redis | Redis | β Yes | Distributed, 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"] }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:
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§π Observability & Tracing
This crate always includes the
tracing crate. The library emits debug and trace span events throughout its internal operations (downloads, cache lookups, subprocess execution, etc.).
β οΈ Important: tracing macros are pure no-ops without a configured subscriber. If you donβt add one, there is zero runtime overhead.
To capture logs, add a subscriber in your application:
[dependencies]
tracing-subscriber = "0.3"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§π Documentation
The documentation is available on docs.rs.
Β§ποΈ 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 extractorsextractor::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
Β§π§© Usage Patterns
- π¬οΈ 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(())
}- π 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§π 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(())
}- π¦ Installing the
yt-dlpbinary 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(())
}- π¦ Installing the
ffmpegbinary 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(())
}- π Updating the
yt-dlpbinary:
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(())
}- π₯ 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(())
}- π 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(())
}- β¨ 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(())
}- π¬ 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(())
}- π΅ 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(())
}- π 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(())
}- βοΈ 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(())
}- πΈ 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(())
}- πΌοΈ 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 1280Γ720
if let Some(thumb) = video.thumbnail_for_size(1280, 720) {
println!("HD thumbnail: {}", thumb.url);
}
Ok(())
}- π 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(())
}- ποΈ 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(())
}- π₯ 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(())
}- π 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(())
}- π 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§ποΈ Format Selection
The library provides a powerful format selection system that allows you to download videos and audio with specific quality and codec preferences.
Β§π¬ Video Quality Options
VideoQuality::Best- Selects the highest quality video format availableVideoQuality::High- Targets 1080p resolutionVideoQuality::Medium- Targets 720p resolutionVideoQuality::Low- Targets 480p resolutionVideoQuality::Worst- Selects the lowest quality video format availableVideoQuality::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)
Β§π΅ Audio Quality Options
AudioQuality::Best- Selects the highest quality audio format availableAudioQuality::High- Targets 192kbps bitrateAudioQuality::Medium- Targets 128kbps bitrateAudioQuality::Low- Targets 96kbps bitrateAudioQuality::Worst- Selects the lowest quality audio format availableAudioQuality::CustomBitrate(u32)- Targets a specific bitrate in kbps (e.g.,CustomBitrate(256)for 256kbps)
Β§ποΈ Codec Preferences
Β§πΉ Video Codecs
VideoCodecPreference::VP9- Prefer VP9 codecVideoCodecPreference::AVC1- Prefer AVC1/H.264 codecVideoCodecPreference::AV1- Prefer AV01/AV1 codecVideoCodecPreference::Custom(String)- Prefer a custom codecVideoCodecPreference::Any- No codec preference
Β§π Audio Codecs
AudioCodecPreference::Opus- Prefer Opus codecAudioCodecPreference::AAC- Prefer AAC codecAudioCodecPreference::MP3- Prefer MP3 codecAudioCodecPreference::Custom(String)- Prefer a custom codecAudioCodecPreference::Any- No codec preference
Β§π§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§π 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.
Β§π§ 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§π 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):
- π 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(())
}- π 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§π₯ Heatmap
Heatmap data (also known as βMost Replayedβ segments) shows viewer engagement across different parts of a video. This feature allows you to identify which segments are most popular:
- π₯ 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(())
}- π 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§π Subtitles
The library provides comprehensive subtitle support, including downloading, language selection, and embedding subtitles into videos:
- π 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(())
}- π₯ 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(())
}- π₯ 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(())
}- π¬ 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(())
}- π 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§π Playlists
The library provides full playlist support, including fetching playlist metadata and downloading videos with various selection options:
- π 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(())
}- π₯ 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(())
}- π― 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(())
}- π 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(())
}- π 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§π Events, Hooks & Webhooks
The library provides a comprehensive event system to monitor download lifecycle and react to events through Rust hooks or HTTP webhooks.
Β§β‘ Event System
All download operations emit events that you can subscribe to:
- π‘ 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§βοΈ Available Events
The library emits 22 different event types covering the entire download lifecycle:
Download Lifecycle:
VideoFetched- Video metadata retrievedDownloadQueued- Download added to queueDownloadStarted- Download beginsDownloadProgress- Progress updates (bytes downloaded, speed, ETA)DownloadPaused/DownloadResumed- Pause/resume eventsDownloadCompleted- Download finished successfullyDownloadFailed- Download failed with errorDownloadCanceled- Download was canceled
Format & Metadata:
FormatSelected- Video/audio format chosenMetadataApplied- Metadata tags writtenChaptersEmbedded- Chapters added to file
Post-Processing:
PostProcessStarted/PostProcessCompleted/PostProcessFailed- FFmpeg operations
Playlist Operations:
PlaylistFetched- Playlist metadata retrievedPlaylistItemStarted/PlaylistItemCompleted/PlaylistItemFailed- Per-item eventsPlaylistCompleted- Entire playlist finished
Advanced:
SegmentStarted/SegmentCompleted- Parallel segment downloads
Β§πͺ Rust Hooks (Feature: hooks)
Register async functions to be called when events occur:
[dependencies]
yt-dlp = { version = "2.7.0", features = ["hooks"] }- π£ Registering a hook for download events:
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 donβt stop downloads)
Β§π 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§π‘ HTTP Webhooks (Feature: webhooks)
Send events to external HTTP endpoints with automatic retry:
[dependencies]
yt-dlp = { version = "2.7.0", features = ["webhooks"] }- π‘ Registering a webhook:
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- π§ Loading a webhook from environment variables:
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
}
}Β§β»οΈ Retry Strategy
- β»οΈ Configuring a retry strategy:
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§π Combining Hooks and Webhooks
Use both hooks and webhooks together:
- π Using hooks and webhooks simultaneously:
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§π 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:
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:
downloadsβ attempted, completed, failed, canceled, total bytes, average speed, peak speed, success ratefetchesβ attempted, succeeded, failed, average duration, success rate (video + playlist fetches)post_processingβ attempted, succeeded, failed, average durationplaylistsβ playlists fetched, failed, per-item success raterecent_downloadsβ bounded history window of completed downloads with per-download details
Β§π Advanced Features
Β§π Proxy Support
The library supports HTTP, HTTPS, and SOCKS5 proxies for both yt-dlp and reqwest downloads:
- π 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§π 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.
Β§Cookie file (Netscape format)
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", β¦
.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§βοΈ 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.
- βοΈ 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(())
}- π 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(())
}- πͺ 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§π΄ 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§π₯ Basic live recording (reqwest engine)
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(())
}Β§π¬ FFmpeg fallback engine
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 stdinq. 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§π‘ Live fragment streaming (Feature: live-streaming)
Enable the feature in your Cargo.toml:
[dependencies]
yt-dlp = { version = "2.7.0", features = ["live-streaming"] }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(())
}Β§π¨ Post-Processing Options
Apply advanced post-processing to videos using FFmpeg:
Β§π§ Basic codec conversion
- π§ 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§ποΈ Advanced post-processing with filters
- ποΈ 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§π 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
Β§β‘ 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§π Available Speed Profiles
Three pre-configured profiles are available:
π’ 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
βοΈ Balanced (for connections 50-500 Mbps) - Default
- 4 concurrent downloads
- 5β20 parallel segments per file
- 8 MB segment size
- 20 MB buffer
- 3 concurrent playlist downloads
- Best for: Most modern internet connections, general use
π Aggressive (for connections > 500 Mbps)
- 6 concurrent downloads
- 6β24 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§π Using Speed Profiles
- π 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§βοΈ Manual Configuration
- βοΈ 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§π 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: Nagleβs algorithm is disabled for lower latency on small writes
- CDN-Friendly Range Probing: Uses
GETwithRange: bytes=0-0instead ofHEADfor 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-dnsfeature 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.
Β§π 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§ποΈ 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+ moreβ¦
For the complete list, see yt-dlpβs supported sites.
Β§π§© Examples
- π¬ 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(())
}- π± 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(())
}- πΈ 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(())
}- π 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§π¬ Profiling (Feature: profiling)
See PROFILING.md for the complete guide (flamegraph, samply, dhat-rs, heaptrack, Criterion, and the raw vs library comparison tool).
Β§ποΈ 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-dlpβs 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 10Results below are averages over 10 runs on a typical broadband connection.
Methodology: raw
yt-dlpre-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.
Β§π΅ Audio streams
| Scenario | yt-dlp | Conservative | Balanced (default) | Aggressive |
|---|---|---|---|---|
| Audio 96 kbps (Low) | 8.26s | 1.27s | 1.47s | 1.27s |
| Audio 128 kbps (Medium) | 7.83s | 1.11s | 1.12s | 1.18s |
| Audio 192 kbps (High) | 8.53s | 1.38s | 1.26s | 1.19s |
| Audio best quality | 8.51s | 1.22s | 1.15s | 1.15s |
Β§π¬ Video streams (no audio)
| Scenario | yt-dlp | Conservative | Balanced (default) | Aggressive |
|---|---|---|---|---|
| Video 480p | 8.59s | 2.39s | 2.11s | 2.58s |
| Video 720p | 9.84s | 3.67s | 3.86s | 3.89s |
| Video 1080p | 17.6s | 8.65s | 8.60s | 8.81s |
| Video best quality | 16.6s | 11.7s | 11.8s | 12.0s |
Β§π¦ 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.
| Scenario | yt-dlp | Conservative | Balanced (default) | Aggressive |
|---|---|---|---|---|
| Native 360p (mp4) | 9.69s | 2.11s | 2.00s | 2.16s |
| Native 720p (mp4) | 20.7s | 2.08s | 2.10s | 2.06s |
Β§π¦ 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).
| Scenario | yt-dlp | Conservative | Balanced (default) | Aggressive |
|---|---|---|---|---|
| Muxed 480p | 9.41s | 3.51s | 3.48s | 3.34s |
| Muxed 720p | 10.4s | 4.95s | 4.92s | 4.97s |
| Muxed 1080p | 18.3s | 10.4s | 10.5s | 10.3s |
| Muxed best quality | 18.2s | 13.7s | 13.7s | 13.3s |
Β§π Speed profiles
| Profile | Parallel segments | Segment size | Use case |
|---|---|---|---|
Conservative | 1β16 | 5 MB | < 50 Mbps connections |
Balanced (default) | 2β20 | 8 MB | Most modern connections |
Aggressive | 3β24 | 10 MB | Fibre / gigabit |
See PROFILING.md for detailed micro-benchmarks.
Β§π‘Features coming soon
-
Use
rust-ffmpegandffmpeg-sys-nextas 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§π 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.
Β§π€ Contributing
Contributions are welcome! Please feel free to submit a Pull Request. Make sure to follow the Contributing Guidelines.
Β§π 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 βyt-dlpβ.
- 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.