← Sticker Wall AGENT · DETAIL

AI Voice Customization & Order Fulfillment System: Solo operator earning 20,000 RMB/month, quietly replicating publishing house audiobook profits

Workflow: Receive customer-submitted materials such as novels and paid knowledge texts daily. First, use text-cleaning tools to au

AGENT

Key Fields

FIELD STAMPS

IndustryMarketing / Advertising

RegionChina（中国大陆）

ScaleSME

ChannelOnline

🔧 Workflow

Receive customer-submitted materials such as novels and paid knowledge texts daily. First, use text-cleaning tools to automatically split chapters and label character dialogues, import them into AI voice cloning tools to batch-generate multi-character audio, manually proofread each section for mispronunciations, adjust emotional phrasing and speaking pace, and finally mix and export MP3-format audiobooks or voiceover files that meet platform requirements. Simultaneously upload them to platforms like Lanren Tingshu, Xiaoyuzhou, and Ximalaya to trigger membership revenue sharing.

🛠 Setup Requirements

Requires a desktop computer equipped with an RTX 3060 or higher GPU for local deployment of open-source voice cloning models, or renting cloud GPU computing power to lower hardware costs. Subscribe to commercial tools such as ElevenLabs for commercial licensing to obtain high-fidelity voice rights, and master basic Adobe Audition mixing skills. The overall setup cycle is about 7 days, with initial test costs under 100 RMB.

🧰 Toolchain

🔧 ElevenLabs
🔧 XTTS-v2
🔧 Adobe Audition
🔧 Wanxiang Voice Platform

💰 Revenue

① Audiobook production outsourcing from publishers and audio platforms (primary revenue): Publishers/platforms pay production fees per book, short books 3,000-5,000 RMB/book, medium-to-long books priced at 15-30 RMB per 1,000 characters × 4-6 books undertaken per month = monthly income of 12,000-25,000 RMB, with monthly income almost entirely coming from this channel (approx. 100%; estimation basis is that this card's monthly income consists of 4-6 delivered books, self-reported case study, independently unverified); ② Content providers billed by episode (replicator tier): 10-minute episodes at 80-150 RMB × 30 short drama episodes = 2,400-4,500 RMB per project, source states 3,000 RMB net profit, accounting for about 20% of monthly income (calculated by dividing 3,000 RMB by this card's median monthly income of 15,000 RMB, media estimate figure, independently unverified); ③ Audio platform playback and subscription sharing: Platforms pay ad revenue sharing based on play counts, CPM 80-150 RMB × single-episode play count 3,000-5,000 views = 240-750 RMB per episode, or subscription-based pricing at ¥12.99/paid episode, ¥299/year paid community. Media estimates this model generates 30,000-50,000 RMB in monthly revenue (media estimate, independently unverified), proportion of total revenue unclear; ④ Opportunity item - Proprietary copyright audiobook matrix: Leveraging batch production costs under 8 RMB per 10,000 characters to lower costs, charging via platform revenue sharing or licensing fees. Both revenue-sharing ratios and licensing pricing are undisclosed, making it impossible to estimate how much incremental value this can bring.

💸 Cost

ElevenLabs commercial subscription is $22 per month, local deployment of the XTTS-v2 model has no subscription fee and only requires cloud computing rental fees of about 0.5 RMB per 10,000 characters, fully automated generation services on platforms like Wanxiang Voice cost under 8 RMB per 10,000 characters, keeping comprehensive production costs per book under 50 RMB.

⏱ Time Investment

Invest 4 to 6 hours daily

🚀 Getting Started

Step one: Post free AI voiceover trial posts in Xiaoshongshu and Douban author communities, contact 3-5 independent authors to offer a trial production of a full AI audiobook under 30,000 characters, grant them free authorization to publish and test on platforms like Xiaoyuzhou. After accumulating 2-3 successful cases, you can connect with publishing house copyright departments or knowledge creator batch voiceover orders.

🔑 Keys to Success

✅ Stable copyright settlement channels with publishing houses or independent authors to ensure a continuous supply of orders
✅ Manual proofreading skills for multi-character voice cloning and emotional pausing to ensure delivery quality
✅ Long-text speech coherence repair capability to avoid character voice drifting, missing words, and mispronunciations
✅ Multi-platform distribution and operation capability to achieve omni-channel compounding monetization of audio content
✅ Commercial voice copyright compliance awareness to avoid platform takedowns and legal claims due to infringement

⚠️ 风险

⚠️ Platforms like Ximalaya and Lanren Tingshu have traffic-limiting policies on purely AI-synthesized voices; quality content must be paired with a small amount of human annotation to gain recommendations
⚠️ Using unauthorized celebrity voice cloning may face risks of platform takedowns and legal claims
⚠️ Long-text generation is prone to missing words, mispronunciations, and character voice drifting issues; heavy rework will lower the actual hourly wage
⚠️ Customer requirements for the emotional subtlety of AI voiceover are high; repeated adjustments will extend the delivery cycle and impact reputation

📌 Real Cases

📌 Wanxiang Voice Platform fully automated AI multi-cast audiobook creation service, production cost under 8 RMB per 10,000 characters, has undertaken batch public-domain book audio adaptation orders from multiple publishing houses
📌 Independent audiobook creator uses the XTTS-v2 + ElevenLabs combination to complete 12 public-domain book AI multi-cast productions in a single month, earning 18,000 RMB per month through membership revenue sharing on Lanren Tingshu and Xiaoyuzhou platforms
📌 AI voice customization team undertakes private publishing house backlist book-to-audio projects, with single-project profits exceeding 3,000 RMB, averaging 5 projects per month and earning over 20,000 RMB monthly

Sources