← Sticker Wall AGENT · DETAIL

AI Voice Customization Specialist: Voice cloning for audiobook production and custom recording, earning over 10,000 RMB per month

Workflow: Every morning, synchronize pending orders from multiple outsourcing platforms, filter text content meeting authorization

AGENT

Key Fields

FIELD STAMPS

IndustryContent / Creator Economy

RegionChina

ScaleSME

ChannelOnline

🔧 Workflow

Every morning, synchronize pending orders from multiple outsourcing platforms, filter text content meeting authorization requirements, use AI tools to clean redundant descriptions, and label tone, stress, and pause nodes for different roles. Input pre-collected legally authorized human voice samples, generate role-specific audio segments through voice cloning tools, batch splice and noise-reduce before exporting the finished product. In the evening, conduct manual emotional coherence calibration, submit for client acceptance, and simultaneously upload to platforms like Himalaya and Lanren Tingshu to trigger long-term revenue sharing. Review client feedback weekly to optimize tone parameters.

🛠 Setup Requirements

Requires basic API calling or local deployment capabilities, acquiring at least 3 legally authorized human voice samples from friends, family, or partners in advance, and training an exclusive tone library to avoid homogenization. For tools, one can self-build an open-source voice cloning tool local environment to reduce long-term costs, or subscribe to commercial voice tools to lower the technical barrier. The overall setup period is about 3-7 days. Running through the first test order allows for officially taking on jobs without professional recording equipment.

🧰 Toolchain

🔧 XTTS-v2
🔧 ElevenLabs
🔧 AI-NovelSpeaker-V2
🔧 Jianying

💰 Revenue

① Audiobook recording fees paid by publishers and authors per book (main income): production fee for a 100,000-word audiobook is 2,000-4,000 RMB × 5-10 orders per month = 10,000-40,000 RMB/month. Mid-tier average is 8,000-12,000 RMB, top-tier is 15,000-30,000 RMB (case study figures, pending independent verification); ② Short drama episodic audio subcontracting (billed per episode): charging 80-150 RMB for a 10-minute single episode, net profit of 3,000 RMB for one 30-episode short drama (case study self-reported figures, pending independent verification), monthly order volume unverified, proportion of total revenue undisclosed; ③ Audio platform playback revenue sharing (authorization sharing): finished products uploaded to platforms share revenue based on play counts, with both share ratios and play counts undisclosed, playback share proportion un-trackable; ④ Opportunity item—maintenance subscriptions and automated mass production: charging bound clients a monthly maintenance fee of 200-500 RMB, binding 10 clients to form stable cash flow (case study figures, pending independent verification), combined with factory-style mass production (cost per 10,000 words under 8 RMB, mass-producing 15-25 books per month, case study figures), maintenance subscription proportion data unavailable.

💸 Cost

The open-source self-built XTTS-v2 solution only requires an annual cloud server fee of about 300 RMB, with no extra fees for audio generation; commercial solutions like the ElevenLabs Pro subscription cost about $22 per month, with additional charges based on generated character volume, bringing the synthesis cost for a single 100,000-word audiobook to under 5 RMB, with no other hardware investment costs.

⏱ Time Investment

3-4 hours per day

🚀 Getting Started

Step 1: Download open-source audiobook generation projects like AI-NovelSpeaker-V2 on GitHub, complete local deployment and test multi-role synthesis effects, or register an ElevenLabs account to familiarize with commercial tool operations; Step 2: Publish 'AI Audiobook Customization/Recording' services on platforms like Xianyu, Zhubajie, and Douban audiobook production groups, setting up a 9.9 RMB / 1,000 words trial package for traffic acquisition; Step 3: After completing the first order, accumulate client case studies, gradually connect with B-end long-term clients such as small and medium-sized publishers, podcasters, and knowledge IPs to form a stable source of orders.

🔑 Keys to Success

✅ Accumulate 15+ compliant exclusive voice libraries covering different genre needs to form a differentiated barrier
✅ Manual calibration eliminates the mechanical feel of AI voices, achieving client repeat purchase rates of over 40%
✅ Connect with 3+ outsourcing platforms to diversify order source risks
✅ Familiarize with audio review rules of platforms like Himalaya and Lanren Tingshu to ensure finished products can be listed for long-term revenue sharing

⚠️ 风险

⚠️ Cloning voices of film/television stars or public figures without authorization from rights holders constitutes infringement
⚠️ Purely AI-synthesized audio judged as non-compliant by platforms like Himalaya and removed
⚠️ Client dissatisfaction with the emotional restoration degree of AI-generated voices leading to refund disputes

📌 Real Cases

📌 Sesame Pie AI Operation: An individual uses XTTS-v2 to undertake novel audiobook recordings, accepting 8 orders a month with an income of 12,000 RMB
📌 Yuanqili Document Case: A single person operates a multi-category audiobook content account, listing on multiple platforms like Lanren Tingshu with monthly revenue sharing exceeding 8,000 RMB
📌 Wanxiang Audio Public Beta Creator: Uses fully automated AI multi-casting tools to batch produce audiobooks, outputting 20 works a month with revenue sharing exceeding 15,000 RMB

Sources