Subscription-on-behalf service: users pay with WeChat Pay/Alipay and WildAI sets up or renews standalone ChatGPT Plus/Pro, Claude Pro/Max, Gemini and X Premium accounts for them, with a full refund if it fails.
Its predecessor was "WildCard (野卡)" (bewildcard.com): in March 2023, right after GPT-4 launched and when users in China could not pay for ChatGPT Plus, it sold overseas virtual credit cards that could be issued "in one minute, no ID needed, RMB supported", and became well known in China's AI circles by riding the demand for paying for ChatGPT. In mid-2024 it stopped issuing cards because of tighter regulation, and in October 2024 switched to "subscribing on behalf of users" (the user pays with WeChat Pay/Alipay and the platform opens ChatGPT Plus/Claude Pro for you); in September-October 2025 it was formally renamed WildAI on the new domain bewild.ai, with "existing users just log in" and balances usable toward subscriptions. So the first users of bewild.ai were WildCard's long-accumulated existing users (the site claims 200K+ users served); new users come mainly from the many "how to subscribe to ChatGPT/Claude" tutorials on Zhihu, Juejin, GitHub gists and personal blogs that carry invite codes (such as AIG, ACCPAY), written by promoters who earn invite rebates.One-line growth story, compiled by us from public sources; speculation is marked separately below.
The main engine is invite rebates: many tutorials on Zhihu, Juejin and GitHub about "how to subscribe to ChatGPT and Claude" carry invite codes, and referrals are the second-largest channel.
The turbo boost is WildCard's existing users migrating over as a group: after the rename, existing users log in directly and their balance can be applied to subscriptions, and they keep coming back to renew monthly, with direct visits at 67%.
By estimate, about 60% of visits are former WildCard users coming back to renew; about 3.4 of every 100 visits click through to the Stripe checkout page.
Model: Lenny's racecar growth framework. Engine = invite rebates (promoters write subscription tutorials with invite codes to earn rebates, and new users come in after reading the tutorials); turbo boost = WildCard's existing users migrating as a group to the new domain; lubricant = monthly renewals, with existing users coming back every month.
Real data: 383,600 visits, month-over-month +11.46%, the largest channel direct at 67.42%, the top-three channel ranking (direct, referrals, organic search), organic search 100% of search, top search terms (wildai, wild ai, bewild, bewild.ai, waildai), top referrers (atatech.org 2.95%, onerway.com 2.72%, zhihu.com 1.92%), social sources (YouTube, X), countries (China 62.22%, Hong Kong 14.34%, Singapore 8.61%, US 8.03%, Taiwan 3.24%), bounce rate 31.34%, 3.9 pages per visit, an average stay of 4 minutes 8 seconds and outbound link destinations onerway.com and stripe.com come from the Similarweb public page for last month; the public page does not give the shares of the second and third channels. 12,894 visits (12,900) to the Stripe checkout page come from quietmoney checkout-page data.
Supporting evidence: WildCard launched in 2023, selling overseas virtual cards that need no ID and support RMB; in 2024 it stopped issuing cards because of tighter regulation and switched to purchasing subscriptions on behalf of users; in 2025 it was renamed WildAI and moved to bewild.ai, with existing users logging in directly; since 2026 promotional posts with invite codes ACCPAY and AIG have appeared on GitHub and Juejin; a user blogged that customer service gave a prorated refund after a failed renewal.
Inference: The two channel circles are real; how each route is split is an inference: direct URL typing 258,600 = former WildCard users 220,000 + typing the URL directly after reading tutorials 23,600 + seeing the comeback news 15,000; other channels 125,000 = tutorials with invite codes 90,000 + seeing the comeback news 20,000 + existing users searching the brand name 15,000. Referrals are second, so the tutorial route is placed mainly in the second circle. The 5% click-through after seeing tutorials and 10% visit rate after seeing the comeback news are guesses. The profile's older channel figures are from an earlier basis and are not used here. The main site mainly pushes payment by WeChat and Alipay, and the Stripe checkout page is only part of it, which is an inference.
Checkout: checkout.stripe.com
“No referral, invite, sharing, leaderboard, watermark or open-source elements”Quote from the site
New-user source: Other (Direct / referral / email / paid). Direct visits 67.4%, referrals 24.4%, search 4.9%; brand keywords 70.4% (wildai, bewild). Referrals come from atatech.org 2.95%, payment-related onerway.com 2.72% and Zhihu 1.92%; the homepage has a "Rewards and Promotion" entry. Returning users come back directly when renewing.
Top channel measured by Similarweb, others estimated by rank
| Attributable new users, last month | Share |
|---|---|
| SEO | 2% |
| Other | 98% |
Direct + brand search 70%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.
| Country | Share of visits |
|---|---|
| China | 62.22% |
| Hong Kong | 14.34% |
| Singapore | 8.61% |
| United States | 8.03% |
| Taiwan | 3.24% |
Traditional Chinese
YouTube, X (Twitter)
28 keywords · brand ~70% · organic 100% / paid 0%
| Similarweb top 5 keywords |
|---|
| wildai |
| wild ai |
| bewild |
| bewild.ai |
| waildai |
Non-brand queries: wildai, waildai
23 referring sites
| Site | Share of referrals |
|---|---|
| atatech.org | 2.95% |
| onerway.com | 2.72% |
| zhihu.com | 1.92% |
| Category | Share of referrals |
|---|---|
| Programming and Developer Software | 84.84% |
| Search Engines | 4.65% |
| Shipping and Logistics | 3.46% |
| Computers Electronics and Technology - Other | 3.21% |
| Social Media Networks | 0.46% |
About 12,894 last month (Similarweb estimate), #309 of 606; the month before was 0. Opening checkout is not paying, so this is not revenue.
New-user source: Other (Direct / referral / email / paid). Mostly referrals, email or paid, or mostly direct with no clear source Direct visits 67.4%, referrals 24.4%, search 4.9%; brand keywords 70.4% (wildai, bewild). Referrals come from atatech.org 2.95%, payment-related onerway.com 2.72% and Zhihu 1.92%; the homepage has a "Rewards and Promotion" entry. Returning users come back directly when renewing.
Exact amount Not found
Dossier confidence: medium. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.
The number: checkout visits is Similarweb's estimate of visits from the site to a checkout payment page; opening checkout is not paying, so this is not revenue. Small sites land on fixed buckets.
Data through last month. Page updated 2026-10-31.