GitHub Acceleration: Full Documentation & Usage Guide
Append the original GitHub link after the acceleration domain to get accelerated access, with no client to install. This guide explains what the five resource types are and what they are used for, with link formats and common tool examples.
Quick start
Paste any GitHub link on the home page and click “Convert” to get an accelerated address, or build one manually in your browser following the rules below.
- Copy the original GitHub link, for example
https://github.com/owner/repo/releases/download/v1.0/app.zip - Prefix it with this site’s domain and a slash, for example
https://ghclone.com/https://github.com/owner/repo/... - Browsers, download tools and git clients can all use the new address directly
Link formats
Three link formats are supported — pick whichever you prefer; the result is identical.
| Format | Example | Notes |
|---|---|---|
| Domain + original link | https://ghclone.com/https://github.com/owner/repo |
Same as mainstream acceleration sites, the most universal |
/gh/ prefix |
https://ghclone.com/gh/github.com/owner/repo |
Saves one https://, a shorter link |
| Bare domain concatenation | https://ghclone.com/raw.githubusercontent.com/owner/repo/main/README.md |
The first segment must be an official GitHub domain |
| Bare repository (clone only) | https://ghclone.com/owner/repo.git |
The shortest address for cloning |
Supported domains include official GitHub domains such as github.com, api.github.com, raw.githubusercontent.com, codeload.github.com, gist.github.com, gist.githubusercontent.com, objects.githubusercontent.com and release-assets.githubusercontent.com.
Five resource types
Download targets on GitHub fall into five categories, each with a different purpose and usage. Skim the overview first, then pick the right approach below.
| Resource type | What it is | Typical use |
|---|---|---|
| Release assets | Artifacts uploaded by the author on the release page, such as installers and build outputs | Install or run software directly, no compiling required |
| Raw files | The verbatim content of a single file in a repository, without the web page wrapper | Run install scripts, read configuration, download single files |
| Source archives | zip / tar.gz generated by GitHub for a branch, tag or commit | Get the full source tree (including branch downloads) without commit history |
| Gists | A lightweight code sharing page, essentially a mini repository | Share small scripts, config snippets and sample code |
| Git clone | Copy the entire repository, including commit history, to your machine with git | Contribute, inspect history, switch branches, build locally |
1. Release assets
For downloading installers from a release page, with resume and multi-threaded download support.
What it is: Artifacts manually uploaded by the author on the Releases page, including compiled binaries, installers, archives and checksum files. These are different from source code: source archives are generated by GitHub automatically, while assets reflect the author’s own release cadence.
What it is for: Most users only need this category — download, then install or run directly without compiling. Assets are usually large and work best with multi-threaded, resumable downloads.
Original: https://github.com/owner/repo/releases/download/v1.0/app.zip Accelerated: https://ghclone.com/https://github.com/owner/repo/releases/download/v1.0/app.zip
2. Raw files
For viewing or downloading a single file from a repository, commonly configuration files and install scripts.
What it is: The verbatim content of a file in a repository, served by raw.githubusercontent.com — the file itself, with no web interface.
What it is for: The most common use is running install scripts (curl ... | bash), reading a single config file, or loading text and data from a fixed address in a program.
Original: https://raw.githubusercontent.com/owner/repo/main/install.sh Accelerated: https://ghclone.com/https://raw.githubusercontent.com/owner/repo/main/install.sh
3. Source archives
zip / tar.gz source archives of a branch or tag.
What it is: A zip / tar.gz that GitHub packages for a branch, tag or commit — the file behind the “Download ZIP” button on the web page; branch downloads belong to this category.
What it is for: Use it when you need the full source tree but do not care about commit history — faster and lighter than cloning. To keep history locally and update continuously, use Git clone below.
Three common targets:
- Branch download:
/archive/refs/heads/<branch>.zip - Tag snapshot:
/archive/refs/tags/<tag>.zip - Commit snapshot:
/archive/<commit-hash>.zip
Original: https://github.com/owner/repo/archive/refs/heads/main.zip Accelerated: https://ghclone.com/https://github.com/owner/repo/archive/refs/heads/main.zip
4. Gists
Both gist pages and raw files are accelerated, and absolute GitHub links inside the page are automatically rewritten to accelerated addresses.
What it is: GitHub’s lightweight code sharing service — a gist is a small, versioned collection of files, essentially a mini repository.
What it is for: Sharing a script, a configuration or an example, usually just dozens of lines that do not warrant a full repository.
Page: https://ghclone.com/https://gist.github.com/owner/gistid Raw: https://ghclone.com/https://gist.githubusercontent.com/owner/gistid/raw/file.txt
5. Git clone & pull
Clone, shallow clone and incremental pull are all supported; only read-only operations are allowed and push requests are rejected.
What it is: Copying the entire repository, with full commit history, to your machine with git — after which you can pull updates, switch branches and inspect the change history.
What it is for: Use it when you need to modify code, run tests locally, build the project, or need the full history; if you only want the source of a specific version, the Archive above is more economical. This site supports read-only operations only — push requests are rejected.
Append -b <branch> to clone a specific branch; use a shallow clone if you only need the latest commit:
git clone https://ghclone.com/owner/repo.git git clone --depth=1 https://ghclone.com/owner/repo.git git clone -b dev --depth=1 https://ghclone.com/owner/repo.git git remote set-url origin https://ghclone.com/owner/repo.git
Common tools
# wget: resumable download wget -c https://ghclone.com/https://github.com/owner/repo/releases/download/v1.0/app.zip # curl: follow redirects and write to file curl -fL -O https://ghclone.com/https://github.com/owner/repo/releases/download/v1.0/app.zip # curl: run an install script directly curl -fsSL https://ghclone.com/https://raw.githubusercontent.com/owner/repo/main/install.sh | bash # aria2: multi-threaded download aria2c -x16 -s16 https://ghclone.com/https://github.com/owner/repo/releases/download/v1.0/app.zip
Releases parsing
Paste a repository home page link on the home page (e.g. https://github.com/owner/repo) and the page parses the repository’s releases, asset sizes and download counts, generating multi-line download buttons for every file.
The parsing capability is also available through its API, convenient for your own scripts and tools:
GET /api/releases?repo=owner/repo GET /api/nodes GET /__probe GET /api/health
FAQ
Prefer lines bound to a custom domain. Services deployed on workers.dev subdomains are not reliably reachable from mainland China, and binding your own domain improves this noticeably. The line probing at the top of the page automatically picks the fastest line.
Parsing relies on the official GitHub API, which is rate-limited per egress IP for anonymous calls (60 requests per hour). Results are cached server-side, and a token can raise the quota to 5,000 requests per hour.
No. This site only proxies public resources and never accepts or forwards any account credentials.
Whether GitHub source archives (archives) support ranged downloads is decided upstream, and some archives return the full content in one response. Release assets generally support resuming.
Simply re-download with a client that supports resuming: wget -c, curl -C - and aria2c -c all continue from where the transfer stopped instead of starting over.
Yes. The page probes the latency of all configured lines concurrently, and converted links and download buttons automatically use the fastest one.
Acceptable use
This service is a download accelerator for public open source resources. It is not a general-purpose network proxy and is not intended to bypass network regulation. Please read the following before use:
- Use the service only for normal access to public open source code, documentation and resources; do not use it for account operations, private repository access, or any purpose that violates laws and regulations
- Make sure your use complies with the laws and regulations of your country and region, and respect the GitHub Terms of Service and the license terms of the original authors
- High-frequency automated requests are prohibited; the server enforces per-IP rate limits
- Account-related paths such as login, registration, session and GraphQL are rejected outright — this is to protect the security of your account
We reserve the right to rate-limit, block or suspend service for violations or abuse without prior notice.