Which Chinese artists are easiest to learn from?
“Try listening to 周杰伦 (Jay Chou)” is somewhat common advice for Chinese learners. It turns out, it’s not great, but not terrible at all for a beginner trying to expand their vocabulary.
I ran 6,545 songs from 200 artists through a word-by-word analysis against the HSK 3.0 lists and ranked them. For each song I segmented the lyrics and measured what share sits in each HSK level, and I plotted the median song for each artist. Here’s what I found.
Most beginner-friendly
| Artist | Typical song in HSK 1–3 |
|---|---|
| Dear Eloise | 81% |
| 陶喆 (David Tao) | 80% |
| 野孩子 (Ye Haizi) | 80% |
| 卢广仲 (Crowd Lu) | 80% |
| TizzyT | 79% |
Least beginner-friendly
| Artist | Typical song in HSK 1–3 |
|---|---|
| 银临 (Yin Lin) | 49% |
| Explosicum | 50% |
| 黄龄 (Isabelle Huang) | 58% |
| 孤独的利里 (Lonely Leary) | 59% |
| 苏紫旭&The Paramecia | 60% |
A 32-point spread. In practice that means at HSK 3, you’d need to look up one word in five for a typical Dear Eloise song, and half of the words in a typical 银临 (Yin Lin) song. A high HSK 1–3 percentage does not mean a song is easy to understand, though — the words beyond HSK 3 are usually the words carrying the image.
Where to start at your level
The graph below ranks artists across two dimensions. The vertical axis is how much of a typical song you already know. The horizontal axis is learnability — how good the words you don’t know are for learning (higher = few unknown words with more repeats).
Showing the 58 best-known artists plus the outliers at each end, out of 152 ranked. An artist needs at least 15 analyzed songs and 2,000 words to be ranked at all. English lines are excluded from the percentages, so rappers are judged on their Chinese alone. No lyrics are stored or shown — these pages are counts only.
The top-right of the chart is the sweet spot: comfortable and packed with a few new words worth learning that are repeated. Slide the scatter to your own HSK level and find the artists that work best for you (song view coming soon). That corner is where you learn the most new Chinese per song, and it is mostly big melodic songwriters:
| Artist | Known at HSK 3 | Learnability |
|---|---|---|
| 野孩子 (Ye Haizi) | 80% | 64 |
| 卢广仲 (Crowd Lu) | 80% | 59 |
| 李荣浩 (Ronghao Li) | 77% | 58 |
| 郭静 (Claire Kuo) | 76% | 58 |
| 王力宏 (Leehom Wang) | 74% | 57 |
李荣浩 (Ronghao Li) is my favorite, so this made me happy and motivated to start memorizing some of his tracks :)
It makes sense that rappers tend to sit in the top-left corner. They score high on raw comprehensibility but low on learnability — their vocabulary is just as simple, but what you don’t already know tends to appear once and never return: 马思唯 (Masiwei) is 76% known yet only 47 learnability score, 艾志恒 (Asen) is 79% known yet only 48, TizzyT is 79% known yet only 48. (English lines are excluded from the percentages, so this is their Chinese alone.)
Gufeng music is the opposite. 银临 (Yin Lin) is one of the two outliers in having high difficulty, alongside Explosicum, and there the genre explains it — Gufeng borrows classical vocabulary and syntax on purpose. It’s beautiful but not good as an on-ramp for the language at my level.
Rock has a wider spread than pop. Beyond and 刺猬 (Hedgehog) sit near the bottom, but 梁博 (Liang Bo) scores pretty high.
In all reality, though, you don’t need to listen to an artist’s “median” song — you can just listen to the easiest ones. For now, you can do that by using Mandoremi; I’ll write another post soon with analysis on individual songs.
What “understandable at your level” actually looks like
Slide the scatter to HSK 1 and the whole cloud drops.

Here is the median share of a typical song a learner at each level already knows, across the 58 artists shown:
| Your level | Median song you understand | Artists where ≥80% is known in the median song |
|---|---|---|
| HSK 1 | 50% | 0 of 58 |
| HSK 2 | 64% | 0 of 58 |
| HSK 3 | 73% | 3 of 58 — Dear Eloise, 陶喆 (David Tao), 野孩子 (Ye Haizi) |
| HSK 4 | 80% | 25 of 58 |
| HSK 5 | 84% | 43 of 58 |
| HSK 6 | 88% | 53 of 58 |
| HSK 7-9 | 95% | 57 of 58 |
The lesson is tough: at HSK 1–2, no artist’s catalogue is comfortable. The most approachable artist at HSK 1, 野孩子 (Ye Haizi), is only 60% understandable there. Three artists save the day for HSK 3 learners — Dear Eloise, 陶喆 (David Tao) and 野孩子 (Ye Haizi) (again though, we’re just looking at the median song). By HSK 5 most artists cross 80%, and by HSK 6 only the genuinely dense writers stay hard — 银临 (Yin Lin) is still at 69%.
HSK split by artist — all 58 artists, ranked
The vocabulary that songs use that HSK doesn’t teach
This was the biggest surprise and is probably worthy of future posts. Two groups stand out: Lyrics really frequently use single-character forms of two-syllable textbook words in HSK 1–5. The single-character version itself is on no HSK list at all. Second, basic romance-related words are super common, but aren’t taught to learners until HSK 7–9.
Single-character word vs its two-character form (the form HSK teaches)
| Single char | Not on HSK | Two-char form | HSK level | Meaning (both) |
|---|---|---|---|---|
| 星 | — | 星星 | HSK 2 | star |
| 身 | — | 身体 | HSK 1 | body |
| 相 | — | 相信 | HSK 2 | (mutual / to believe) |
| 影 | — | 电影 | HSK 1 | (shadow / movie) |
| 何 | — | 任何 | HSK 3 | (what / any) |
The emotion words — taught, but only at HSK 7–9
| Word | HSK level | Meaning | In how many songs |
|---|---|---|---|
| 情 | HSK 7–9 | love / feeling | 10.8% of songs |
| 寂寞 | HSK 7–9 | lonely | 8.6% of songs |
| 温柔 | HSK 7–9 | tender / gentle | 8.7% of songs |
| 思念 | HSK 7–9 | to miss someone | 6.0% of songs |
| 孤单 | HSK 7–9 | lone / solitary | 5.1% of songs |
| 吻 | HSK 7–9 | kiss | 5.3% of songs |
More on this coming soon!
Now you try!

Want to analyze your own playlist, or check a song? I built this tool because I wanted this for my own playlist. You paste a song’s lyrics or import a playlist and it scores everything against your actual level and your own known-words list, for free. For now, you need to source the lyrics yourself and copy-paste them onto this site — but if someone else has already checked that song, including the 6.5k that went into this analysis, the HSK analysis will already be done to point you in the right direction after a big playlist import. I’m also happy to run the numbers on big artists if I’m missing any, and more analysis will be coming in future posts!
Figures fixed on 2026-09-02, from 6,545 analyzed songs.