Published April 9, 2026 by Soma Yano, Japanese software engineer and sole operator of Petarico. This page is written in a mix of Hinglish and English because that is how most Indian mobile searches for tech queries are typed — if you prefer pure English or pure Hindi, apologies, I only speak Japanese and English fluently and have kept the technical English crisp rather than attempting a Hindi translation I cannot verify.
For impatient readers: just paste the URL below, the tool works. The rest of this page is for readers who want to understand exactly what is happening on the server, what Indian law actually says, and why I built this.
The tool is directly below. Paste a YouTube link into it and you will get an MP3 file back in tens of seconds. No signup, no app, no ads, no tracking. The rest of this page explains, in plain terms, how the tool works, what it will not do, where the code runs, and what Indian law actually says about personal downloads.
Paste YouTube URL here — get a 128 kbps MP3 in ~30 seconds
India's mobile data is cheap by global standards, but most prepaid plans have a hard monthly cap. Rs 26 = 2 GB for 28 days. Rs 62 = 6 GB. The numbers below are not opinions; they come from the fact that a 4-minute song is about 3.75 MB as a 128 kbps MP3, and about 24 MB when streamed at standard quality. If you replay the song ten times, streaming costs ten times more data than downloading once. Change the sliders to see how this maps to your plan.
See how much Jio data you save by downloading songs once instead of streaming them repeatedly. All numbers derived from public specs (128 kbps MP3 = ~3.75 MB for a 4-minute song; standard streaming ≈ 40 MB/hour).
Open the song on YouTube in any browser or the YouTube app. Tap Share, then Copy link. On desktop, copy the URL from the address bar. Both the short form (https://youtu.be/...) and the long form (https://youtube.com/watch?v=...) work.
Paste the link into the input. The tool will automatically detect the video, show you the thumbnail, and ask which format you want. Choose MP3.
A progress bar updates in real time, one percent at a time, so you can see exactly what is happening. Conversion time varies with video length: typical three-to-five-minute songs complete in about 10-20 seconds, while a full DVD or concert recording can take up to a minute or so. When the bar reaches 100%, a Download button appears. Tap it, and the MP3 is saved to your phone's Downloads folder with the correct song name as the filename (not video_xH3kj.mp3 like many other sites).
My name is Soma Yano. I am a Japanese software engineer, graduated from Shizuoka University (Computer Science) in 2023, now working full-time as an engineer at HCLTech after two years at Hewlett Packard Japan. Petarico is my personal side project — no VC money, no team, no co-founder, just me paying the AWS bill from my salary.
The origin story is mundane: before an international flight I wanted to save a few YouTube videos to my phone to watch offline. Every "free mp3 download" site I tried was covered in aggressive pop-ups, several crashed in the middle of a download, and two redirected me to casino landing pages. I thought "this is a simple need, there should be a clean browser-based tool", and when I could not find one I built it.
India is the largest YouTube music audience in the world — over 550 million monthly active music listeners, according to Music Ally's 2025 market report. At the same time, the paid music streaming market in India is tiny: Spotify India has about 3 million paid users and Gaana about 1.4 million, against roughly 550 million YouTube music users. That means fewer than one in a hundred Indian music listeners pay a monthly streaming subscription. The rest listen for free, which for most apps means with advertisements and without offline playback.
At the same time, Jio Phone (the Rs 999 feature phone) and older Android devices cannot install modern streaming apps without running out of memory. A Rs 26 Jio plan gives 2 GB for 28 days. If you stream a 4-minute song on standard quality, that is about 24 MB — one song repeated ten times eats a day of data. Downloading the same song once and replaying it locally costs about 3.75 MB and zero further data.
I am not an expert on Indian music culture. I do not know the difference between a good Sidhu Moose Wala track and a great one. I cannot tell you which Pawan Singh song is the anthem of Bhojpuri weddings this year. What I can do is keep the tool working, keep it honest, keep it free, and explain — as an engineer — exactly what happens when you press the button. That is what the rest of this page is about.
Petarico's conversion pipeline lives in AWS and has five stages. Reading this section will not make the tool work any better — but it will let you understand exactly where your data goes and why certain things are impossible. If you are a developer, you might appreciate the fact that the entire architecture is boringly standard rather than a stack of mystery black-boxes.
Your browser sends the YouTube URL over HTTPS to an AWS API Gateway endpoint in the Tokyo region. There is no CORS trickery, no proxy obfuscation, no hidden WebSocket. The request contains the URL, the requested format (MP3 or MP4), and a randomly generated browser ID that exists only to track your job progress. No email, no IP address logging, no cookies.
The request hits an AWS Lambda function running a Python FastAPI app. The function validates the URL (must be a public YouTube video, under 2 hours long), creates a job record in DynamoDB with status queued, and pushes a message to an SQS queue. Then it returns a job ID to your browser. This step takes about 200 milliseconds.
A separate worker Lambda, running as a Docker container with yt-dlp and ffmpeg pre-installed, picks the message off the queue. yt-dlp (a fork of youtube-dl, the de facto industry standard) fetches the video's audio stream from YouTube over an authenticated European proxy — Germany, Italy, France, or United Kingdom, chosen randomly from a pool of 15 IPs. This is the step that takes the most time, usually 5 to 30 seconds depending on the video length.
The audio stream is decoded and re-encoded to a constant-bitrate 128 kbps MP3 by ffmpeg using the LAME encoder. ffmpeg is the tool that runs behind the scenes at every major video service on the internet, including YouTube, Netflix, and Instagram. The output is a plain MP3 file that any music player or DJ mixer can read.
The MP3 is uploaded to an Amazon S3 bucket in Tokyo. S3 generates a pre-signed URL that is valid for one hour. The URL is written back to the job record in DynamoDB, and the job status becomes done. A lifecycle rule on the bucket automatically deletes every file after one hour, so nothing accumulates on Petarico's servers.
Meanwhile, your browser has been polling the API for the job status every second. When the status becomes done, the API returns the pre-signed URL. Your browser follows the URL and pulls the MP3 directly from S3. This last step runs at your local Jio 4G speed, and is usually over in 1-5 seconds for a 4 MB file.
Double-blind listening tests published in audio engineering journals (notably the ones from the Audio Engineering Society) show that for typical mobile listening — phone speaker, in-ear earphones up to roughly Rs 2,000 — the vast majority of listeners cannot reliably distinguish 128 kbps MP3 from uncompressed CD audio. The ear-brain system simply cannot hear the compression artifacts at that listening volume and on that hardware. Upping the bitrate to 320 kbps would double the file size (and double the Jio data you spend when downloading) without any perceptible improvement on the devices most Indians use.
If you are a studio engineer with monitor speakers and trained ears, you should not use Petarico — you should buy the lossless master from the label. For everyone else, 128 kbps is the honest sweet spot.
Apps are a trap for Indian mobile users. They require installation (which needs free storage on the device), they need Play Store distribution (which Google removes regularly for "mp3 download" apps), they fragment across Android versions, they cannot be used on feature phones, and they accumulate background services that drain the battery. A website has none of these problems. It opens, it works, you close it. The only requirement is a browser updated to 2020 or later — which is true of almost every phone sold in India in the last six years.
Being honest about limitations is a better use of your time than listing only the good things. Here is what this tool cannot or will not do:
I am a software engineer, not a lawyer. I cannot give you legal advice, and if you need a definitive answer for a commercial use case (wedding DJ, event, broadcasting, restaurant playlist) you should consult a qualified copyright lawyer in your state. What I can do is link you to the actual text of the law and to independent commentaries so you can read it yourself.
This is the section of Indian law that covers "fair dealing". It says that fair dealing with any work, "for the purposes of private or personal use, including research", does not constitute an infringement of copyright. The majority legal interpretation is that downloading a song for your own personal, non-commercial listening falls within this exception. You can read the full text on the government's official code repository:
→ indiacode.nic.in — Copyright Act 1957 Section 52 (official)
The fair dealing doctrine in Indian copyright law has been the subject of extensive legal analysis. Three readable commentaries from independent law firms and academic sources:
India has two music copyright societies. The Indian Performing Right Society (IPRS) collects royalties on behalf of composers, lyricists, and music publishers for public performance and electronic transmission of musical works. PPL India collects royalties for sound recordings on behalf of record labels. Neither society's license is required for personal, non-commercial use of a downloaded file inside your home or on your phone. Both licenses become relevant when music is played in a public or commercial setting — restaurant, wedding venue, shop, public event. If you are a DJ, event organizer, or venue owner, talk to a lawyer about IPRS and PPL licensing.
Petarico is not the right tool for everything. Here are four alternatives and the exact situations in which each is genuinely better.
If you use Jio and mostly want to listen to Hindi and regional film music, JioSaavn Free is an excellent first option. Jio SIM holders get 30 days of JioSaavn Pro (ad-free with offline downloads) bundled free. You get a huge catalog, personalized playlists, discovery, and no need to paste URLs. The trade-off is that JioSaavn downloads are encrypted and locked inside the JioSaavn app — they cannot be copied to a PA system, USB stick, old car stereo, or DJ mixer. Use JioSaavn for discovery and casual listening; use Petarico when you need a portable MP3 file.
If you already use YouTube Music heavily and want to remove ads and enable offline playback inside the YouTube Music app, the official Premium subscription is the cleanest option. The offline downloads, though, remain inside the app — you cannot move the files. If that is acceptable and you can afford Rs 129 a month, YouTube Music Premium is a better fit than Petarico for heavy users. If the monthly fee is a problem, or if you need the file outside the app, use Petarico.
Gaana has a strong catalog of Bollywood, regional, and devotional music, and Gaana Plus removes ads and enables unlimited MP3 downloads inside the Gaana app. The app-only limitation is the same as JioSaavn — you cannot copy the files out. If you want a well-curated Indian-first streaming experience and don't need file portability, Gaana Plus is a solid choice.
Spotify has the best algorithmic playlists and the largest international catalog. Spotify Individual is Rs 119/month in India. The free tier does not allow offline playback. If you primarily listen to international music or want Discover Weekly and similar playlist personalization, Spotify is unbeatable. If you mostly listen to Indian music, JioSaavn or Gaana may serve you better, and if you need a portable MP3 file, Petarico is the only free option.
At the top of this page there is a panel showing Petarico's actual statistics for the last seven days — total jobs, success rate, average conversion time, and the size of the primary and fallback proxy pools. These numbers are not marketing estimates. They are pulled directly from the same database that runs the tool, via a public HTTP endpoint that you can call yourself:
→ /v1/public-stats (public JSON endpoint)
As of the publication of this article ( 11 April 2026) the numbers are:
Let me be completely honest about the scale: Petarico is a one-person side project and the user base is small. Nineteen jobs in a week is not much — and growing. This page is the first time I am actively inviting Indian users, so the numbers above represent the pre-launch baseline. If you are reading this in the weeks after publication, the numbers should already be higher (the live panel at the top of the page always shows the most recent 7-day window). I want the tool to serve more people who actually need it — Indian users on Jio plans, people on feature phones, sound technicians at small-venue events. As the user base grows, I will update the Changelog below and leave these older numbers visible so that growth is traceable.
If these numbers ever look suspicious or stale, call the endpoint above yourself and compare. The methodology is:
__health_check__) excluded — those are automated tests, not real user trafficelapsed_sec for done jobs onlyIn the spirit of open-source software, I keep a public log of every meaningful change to Petarico's infrastructure and this page. If you spot an inaccuracy, the Changelog is how you see when and whether it was fixed.
/v1/public-stats honest transparency endpoint showing cumulative stats since launch.NONE placeholder in configuration) and all traffic was going through the US fallback, which was causing YouTube bot detection. Fixed by setting the 15-IP EU pool (Germany, Italy, France, UK) as the primary.This page will be updated at least once per quarter with fresh backend statistics and any infrastructure changes. Next scheduled update: July 2026.

Soma Yano — Japanese software engineer based in Tokyo. BSc in Computer Science from Shizuoka University (2023), currently a full-time software engineer at HCLTech. Petarico is my personal side project, maintained alone on weekends and evenings, with the AWS bill paid from my salary.
The majority interpretation of the Copyright Act 1957, Section 52(1)(a) is that personal, non-commercial downloads fall within fair dealing. Commercial use, redistribution, and public performance are separate questions. For commercial use cases, consult a lawyer. Primary source and independent commentaries are linked in the legal section above.
Yes. Petarico is a website, not an app. The browser that ships with JioPhone and Jio Bharat can open petarico.com and use the tool normally. There is no installation, no permission prompt, and no background service. The interface may look simpler on a small screen but the functionality is the same.
Because double-blind listening tests show that 128 kbps is indistinguishable from higher bitrates on typical mobile hardware (phone speakers, in-ear earphones up to about Rs 2,000). Going higher doubles the file size, which doubles the Jio data cost, with no audible benefit. If you need lossless audio for a studio monitor system, buy the lossless master from the label rather than using any free tool.
Files are kept on S3 only long enough to generate a download URL, then deleted within an hour by a lifecycle rule. Petarico uses Google Analytics to understand aggregate site usage, but there is no Facebook Pixel, no third-party tracking cookies beyond GA, and no user accounts. The only other logs are aggregate counters (used to populate the public stats endpoint) that contain no personal data.
The most common cause of failure is a video that is private, unlisted, geo-restricted outside Europe, or longer than two hours. If none of those apply and the tool still fails, please email the details to petarico.labs@gmail.com — include the YouTube URL, the time of day, and the device you were using. I read every one personally because Petarico is a one-person project.