parth-dl is a fast, open-source Instagram Reels, Posts, and Profile Picture Downloader built entirely in Python. Use it three ways — a command-line tool, a Python library, or a local HTTP API that any app in any language can call.
🧠 Designed for automation • ⚙️ Developer-friendly • 🪶 Zero Dependencies • 🔐 Secure-by-default
Keywords: instagram downloader, python instagram downloader, reels downloader, instagram scraper, insta video downloader, parth-dl, parthmax, saksham pathak, python cli, open source downloader
Built by @parthmax_ — Minimal, Fast, and Developer-Centric.
pip install parth-dl# Download a reel (with audio)
parth-dl https://www.instagram.com/reel/Cxyz123AbCd/
# Keep going — prompts for the next URL after each download
parth-dl -i
# Open the web UI in your browser
parth-dl servefrom parth_dl import download
download("https://www.instagram.com/reel/Cxyz123AbCd/")
Download Instagram reels, posts, or profile pictures — directly from your terminal.
| Use when | Docs | |
|---|---|---|
| CLI | Terminal, shell scripts, CI. --json + typed exit codes make it scriptable. |
docs/cli.md |
| Python library | You're in Python. from parth_dl import download, get_info. |
docs/python-api.md |
| HTTP API | You're not in Python. parth-dl serve → JSON over localhost, for Node, Go, PHP, a browser… |
docs/http-api.md |
The dict every surface returns is documented field-by-field in docs/schema.md — that's the contract to build on. Copy-paste integrations for Node, Go, PHP, Discord and Bash are in docs/recipes.md.
- ✅ Reels — full video with synchronized audio
- ✅ Posts — single images, videos, and mixed carousels
- ✅ Profile pictures — HD, no login
- ✅ Resumable downloads — interrupted transfers pick up where they left off; a partial file is never mistaken for a complete one
- ✅ Clickable results — finished files print as terminal hyperlinks that open in your player
- ✅ Interactive mode —
-ikeeps prompting for the next URL - ✅ Web UI —
parth-dl servefor a paste-a-link interface with a live progress bar and a download queue - ✅ Zero dependencies — pure Python standard library, no supply-chain risk
- ✅ Smart rate limiting — 30 req/min sliding window, avoids IP bans
- ✅ Jittered exponential backoff — resilient to transient network failures
- ✅ Cross-platform — Windows, macOS, Linux; Python 3.8+
parth-dl serveOpens a local page at http://127.0.0.1:8000: paste a link, see a preview, watch a real
progress bar, click to save. Queue up as many links as you like — they download in the
background while you paste the next one.
It's loopback-only, has no CORS header, and rejects non-loopback Host headers, so no
website you visit can drive your downloader. See docs/http-api.md.
parth_dl/
├── __init__.py # Public API: download(), get_info(), InstagramDownloader
├── core.py # Orchestration: extraction -> format selection -> resumable transfer
├── extractors.py # Multi-layer extraction: Embed -> GraphQL -> API, with fallback
├── utils.py # Retry, rate limiting, sanitization, validation, progress
├── cli.py # argparse CLI
├── server.py # `parth-dl serve` — JSON API + web UI
└── web/index.html # Single-file UI (works served, or static as a landing page)
Multi-layer fallback extraction. Instagram will happily 403 one endpoint while another still answers, so every extraction method gets its turn before parth-dl gives up.
- Zero dependencies — nothing but the standard library, no supply-chain surface
- URL validation — media is only ever fetched from Instagram's own CDN hosts, never an arbitrary URL
- Filename sanitization — path traversal, Windows reserved names, control characters
- Timeout-safe requests — no hangs
- Loopback-only server — with DNS-rebinding protection
For educational, personal, and research use. You are solely responsible for complying with Instagram's Terms of Service and applicable law.
git clone https://github.com/parthmax2/parth-dl.git
cd parth-dl
pip install -e ".[dev]"
pytest # 91 tests
ruff check .Tests run against a real loopback HTTP server rather than mocks, so resume, Range
handling, truncation and retry behaviour are exercised for real.
Contributions and PRs welcome — fork, branch, commit, open a PR. If you build something
with parth-dl, tag @parthmax_.
MIT — see LICENSE. Provided "as is", without warranty.
Saksham Pathak (Parthmax) — Generative AI Engineer | Python Developer | Open Source Creator
- 🌐 parthmax.in
- 🧑💻 GitHub
- 🤗 Hugging Face
- FALCON — Fake News Analysis & Verification System
- FaceAging AI — AI-based Face Aging & Transformation
- 100 Best GenAI Projects — Curated open-source index
⭐ Star the repo if you find it useful 💡 Built with ❤️ and Python by Parthmax