Generating Chinese songs with overseas models like Suno often produces a distinctly "foreign accent" — muddy pronunciation, scrambled tones, and rap sections that are especially awkward. GoGo AI (Gege), from Hangzhou Yinli Shandong, was built to solve exactly this: it is China's first pure-Chinese AI music foundation model, optimized for Chinese-language songwriting, generating a complete Chinese song in 10 seconds on a single GPU, and it is already integrated with seven ByteDance-family platforms including Douyin and CapCut. This piece breaks down its capabilities, how to access it, and how creators can put it to work.

What GoGo AI Is

GoGo AI is a billion-parameter-class AI music foundation model developed by Hangzhou Yinli Shandong Artificial Intelligence Technology Co. (founder Long Yong), positioned as "the AI music model that understands Chinese best."

Its core differentiation:

• Pure-Chinese optimization: Mandarin enunciation, tones, and Chinese-instrument arrangement were trained specifically, clearly outperforming overseas models
• Fast generation: a complete song in 10 seconds on a single GPU
• End-to-end integration with the ByteDance platform family: a closed loop from generation to distribution to monetization

What GoGo AI solves is not "can it generate music" but "does the generated Chinese song sound like it was sung by a native speaker."

Core Capabilities

Mandarin enunciation and emotion

The biggest selling point is authentic Chinese pronunciation. Overseas models (Suno, Udio) commonly struggle with Chinese lyrics:

• Flat/retroflex confusion (z/zh, s/sh, c/ch)
• Misplaced tones (singing "gushi" (story) as "gushi" (attend to matters))
• Traditional Chinese instruments faked with electronic synthesizers

GoGo AI is optimized specifically along these dimensions; according to third-party evaluations, its generated Chinese songs are clearly better than overseas models on naturalness and enunciation accuracy.

Multi-genre coverage

It supports mainstream Chinese genres:

• Pop, folk, rap (Chinese flow is a highlight)
• Guofeng, ancient-style, and traditional Chinese music (real samples of erhu, guzheng, flute, etc.)
• Electronic, rock, R&B

Generation speed

A complete song (vocals + arrangement) in 10 seconds on a single GPU — compared with peers that routinely take tens of seconds, it fits real-time interactive scenarios (live-stream song requests, embedded in content-creation tools).

Seven-Platform Integration: a Monetization Loop for Creators

GoGo AI is already wired into seven ByteDance-family platforms, which is what separates it from pure model vendors:

1. Douyin — short-video BGM
2. CapCut (Jianying) — video-editing soundtracks
3. Soda Music — music streaming
4. Tomato Novel Audio — audiobooks / music
5. Xigua Video — mid- and long-form video BGM
6. Toutiao — article/video soundtracks
7. Other ByteDance-family content platforms

The Monetization Path

For creators, the loop looks roughly like this:

1. Generate an original Chinese song on GoGo AI (with full copyright)
2. The song is listed on Soda Music and the Douyin music library
3. Creators (including you) use these songs as video BGM in Douyin and CapCut
4. The song earns membership revenue and ad revenue on Soda Music and similar platforms
5. Revenue is shared proportionally with the original AI creator

That means the path of an individual writing songs with AI → fully compliant listing across the ByteDance family → collecting revenue share is already open for business — no more "generated it, now nowhere to post it."

How to Use It

Entry Points

• Trial on the official platform (search "GoGo AI" or "Gege")
• Indirect access via the built-in AI music features in Douyin / CapCut

Typical Creator Workflow

Step 1: Write lyrics or pick a theme

You can paste in lyrics directly, or just describe a theme ("a folk song about a summer beach"), and the model fills in the rest.

Step 2: Pick a genre

Choose from presets like pop, folk, guofeng, and rap; some platforms support a reference audio clip.

Step 3: Generate and listen

A demo comes back within 10 seconds; if it's not right, regenerate or fine-tune parameters (vocal timbre, tempo).

Step 4: Publish and distribute

Once you're satisfied, publish through the platform tools to Soda Music and other ByteDance-family platforms, and it automatically enters the Douyin music library.

Tip: before publishing, confirm your lyrics have no copyright issues (tracks using someone else's lyrics will fail review).

Best-Fit Scenarios

Common Issues

• Who owns the copyright to generated songs: usually the creator, but confirm the platform agreement before commercial use; terms may differ across plans.
• Can I publish to Spotify / Apple Music: GoGo AI currently integrates mainly with the ByteDance family; overseas distribution is on you.
• The rap sections still come out mushy: rap flow is hard for every AI music model; cut fast passages shorter and slow the tempo, and results improve a lot.
• Enterprise API access: contact Hangzhou Yinli Shandong's business team for model API integration into your own product.

---

Related Links

• BAAI Community coverage: https://hub.baai.ac.cn/view/56296
• ByteDance-family platforms: Douyin / CapCut / Soda Music