💡 AI 模型在英文提示詞下表現最佳。因此,提示詞本文以英文呈現。使用英文輸入可獲得更準確、更詳細的回應。 你的每週推薦已經乏味,最愛的歌也聽到膩了。這些提示詞讓 AI 變成真正懂「氛圍」而非只懂曲風的音樂策展人。經過四大 AI 模型實測,找出哪個最有音樂品味。
提示詞
依照氛圍建立播放清單
I need a playlist for this vibe: [Describe the feeling, e.g., 'late night drive through a rainy city' or 'Sunday morning making pancakes'] Preferences: - Genres I like: [list genres] - Artists I already love: [list 3-5 artists] - Vocal preference: [vocals / instrumental / mix] - Era: [modern only / classics ok / any decade] Give me: 1. A playlist name that captures the vibe 2. 15 songs in the ideal listening order 3. For each: artist, song title, year, and why it fits the vibe 4. 3 deep cuts most people haven't heard 5. The BPM arc of the playlist (how energy flows) 6. A Spotify/Apple Music search string for each track
進階技巧
描述場景,而非曲風。「為約會對象做晚餐」比「放鬆的獨立音樂播放清單」能得到好得多的結果,因為 AI 能將能量、節奏和氛圍與場景配對。
已測試 Mar 15, 2026
打造運動配對播放清單
Build me a workout playlist for: Workout type: [running / lifting / HIIT / yoga / cycling] Duration: [30 min / 45 min / 60 min] Intensity pattern: [steady / intervals / build-to-peak] Genres I like: [list genres] Artists I'm tired of: [list overplayed artists] Give me: 1. Songs matched to each workout phase (warmup, peak, cooldown) 2. BPM for each track and why it matches that phase 3. Total playlist duration matching my workout length 4. 3 tracks specifically for the hardest interval 5. A motivational "power song" for when I want to quit 6. Alternative tracks if I don't like a pick
進階技巧
告訴 AI 你的跑步配速或重訓節奏。160 BPM 的歌非常適合快跑,但對深蹲來說太快了。讓 BPM 與動作速度匹配,差異非常大。
已測試 Mar 15, 2026
打造提升專注力的播放清單
I need music to help me focus on [type of work: writing, coding, studying, design]. Session length: [25 min pomodoro / 2 hour deep work / all day] Noise preference: [silence-adjacent / ambient / rhythmic / lo-fi] Lyrics: [no lyrics / foreign language ok / some lyrics fine] What distracts me: [sudden changes, heavy bass, vocals, etc.] Give me: 1. 15 tracks designed for sustained focus 2. For each: artist, title, and why it aids concentration 3. The science behind why this type of music helps focus 4. A structure: opening tracks to settle in, core focus tracks, gentle wind-down 5. 3 ambient/soundscape options as alternatives 6. One playlist to avoid for focus work and why
進階技巧
說明什麼會讓你分心,而非只說你喜歡什麼。如果突然的節奏變化會打斷你的專注,AI 會避開動態變化大的曲目——但你得先告訴它。
已測試 Mar 15, 2026
打造人人都愛的派對播放清單
I'm hosting a [type of event: house party, dinner party, BBQ, birthday]. Guest profile: [age range, music taste diversity] Vibe progression: [background → dance floor / steady chill / building energy] Duration: [3 hours / 5 hours / all night] Must-play requests: [any songs guests specifically asked for] Avoid: [genres or artists that won't work for this crowd] Build my party playlist: 1. 30+ tracks organized by phase (arrival, dinner, dance, wind-down) 2. For each phase: the energy level and why these tracks fit 3. 5 "universal crowd-pleasers" everyone knows 4. 5 tracks that will make people say "oh I love this song!" 5. Transition tracks between phases so energy shifts feel natural 6. A "rescue song" to play if the vibe drops
進階技巧
先列出賓客的必播歌曲。AI 會以這些歌為核心打造整份播放清單,創造出連貫的曲目流,而非把點播硬塞進隨機清單裡。
已測試 Mar 15, 2026
發現新的藝人和曲風
I want to expand my music taste. Here's what I currently listen to: [List 5 favorite artists and what you love about each] I'm curious about: [genres or scenes you want to explore] I've tried and bounced off: [genres that didn't click] Open to: [international music / indie / experimental / mainstream] Map my musical expansion: 1. 5 artists I'd love based on my current taste but haven't heard 2. A "gateway album" for each new genre I want to explore 3. 3 adjacent genres I didn't mention but would probably enjoy 4. A listening path: which artist to try first and in what order 5. One genre I dismissed that deserves a second chance (and the right entry point) 6. Music podcasts or channels that match my taste evolution
進階技巧
說明你喜歡每位藝人的「哪個部分」,而非只列名字。「我喜歡 Radiohead 的實驗性製作」和「我喜歡 Radiohead 的情感歌詞」會得到完全不同的推薦。
已測試 Mar 15, 2026
基於實際測試結果 — 非假設推測。 查看測試方法
Claude Sonnet 4
最擅長理解抽象氛圍和情緒曲線。打造的播放清單像是有品味的人親手挑選,而非演算法組裝。
最佳氛圍配對Gemini 2.5 Pro
跨年代和曲風的音樂知識最廣博。擅長跨時代的連結和音樂脈絡的梳理。
最佳音樂知識GPT-4.1
最適合運動、派對和需要能量管理的場合。比其他模型更懂群眾動態和 BPM 配對。
最佳功能性播放清單Grok 3
提供跨曲風的推薦,對新興藝人和地下音樂場景有敏銳的嗅覺。
最佳新興藝人描述場景,而非曲風——「日落時分開著車窗行駛在海岸公路上」比「獨立搖滾」建立的播放清單好上太多
提供 3-5 首定調曲目,讓 AI 能反推出共同元素
要求歌曲之間的銜接流暢度,而非只要歌曲本身——好的播放清單有起承轉合的曲線