¥399 every 4 weeks for access to GPT-4o and Claude 3.5 Sonnet — that’s the core proposition at aigocode.com, and it’s a very different pricing model from the per-token billing most Chinese developers are used to.
Pricing Breakdown
Aigocode.com doesn’t use per-million-token rates. Instead, you pay a flat ¥399 subscription every 4 weeks. There’s no free trial, no pay-as-you-go option, and no hidden per-token fees.
| Plan | Price | Duration | Models Available | Notes |
|---|---|---|---|---|
| Monthly Subscription | ¥399 | 4 weeks | GPT-4o, Claude 3.5 Sonnet | No per-token billing |
| Pay-as-you-go | Not available | N/A | N/A | Subscription only |
| Free tier | None | N/A | N/A | No free trial |
| The math: ¥100/week for roughly ¥110 worth of credit. That’s a 10% premium on face value — you’re paying more than the credit you receive. Compare this to direct API access where GPT-4o costs roughly ¥0.03 per 1K input tokens via OpenAI’s official pricing (assuming no VPN costs). If you’re generating less than ~3.3 million input tokens per month on GPT-4o, the subscription likely costs more. |
Pros & Cons
Pros:
- No channel mixing — testing showed clean routing to actual provider models
- Supports both Alipay and WeChat Pay, which is standard for Chinese users
- Stable during test period — no dropped requests or timeouts observed
- Flat pricing eliminates token-count anxiety; you know exactly what you’ll pay Cons:
- Expensive relative to credit value — ¥100/week for ¥110 credit means you’re paying a premium
- Model selection is thin: only GPT-4o and Claude 3.5 Sonnet. No Gemini, no DeepSeek, no smaller/cheaper models
- 94% uptime is below the 99%+ standard for production API relay services
- No refund policy specified — if you pay ¥399 and the service degrades, you’re locked in
- Occasional stability fluctuations reported, which is concerning at this price point
Cost Comparison vs Direct API
Direct GPT-4o access via official API (assuming you have a payment method): $5 per 1M input tokens, $15 per 1M output tokens. For ¥399 ($55 USD), you’d get roughly 11M input tokens or 3.6M output tokens.
Aigocode.com’s ¥399 gives you ¥110 credit equivalent — that’s about $15 USD worth of tokens. You’re paying 3.6x the direct API price for the convenience of no VPN and Chinese payment methods.
For light users (under 500K tokens/month), the subscription is clearly worse. For heavy users generating 5M+ input tokens monthly, the subscription might break even — but only if you exclusively use GPT-4o and Claude 3.5 Sonnet.
Verdict
Aigocode.com works as advertised — no channel mixing, stable connections, and Chinese payment support. But the pricing model is backwards for most developers. ¥399/month for two models at a 10% premium over credit value, with 94% uptime and no refund policy, is hard to justify. If you’re a heavy GPT-4o/Claude 3.5 Sonnet user who hates tracking token counts, this might fit. Everyone else should look for per-token relay services with better model selection and uptime guarantees.
FAQ
Q: Can I try aigocode.com before paying ¥399? A: No. There is no free trial and no pay-as-you-go option. You must commit to the full subscription. Q: What happens if I use up my ¥110 credit before the 4 weeks end? A: The platform data doesn’t specify. There’s no mention of top-up options or overage billing. You may need to wait for the next billing cycle. Q: Does aigocode.com support DeepSeek or Gemini models? A: No. The platform explicitly lists only GPT-4o and Claude 3.5 Sonnet. Despite the “Pros” mentioning Gemini support in the data, the Models field only shows two providers.
Data provenance: any figures from hands-on checks are author-reported and tested on the recorded date. Independent verification is unavailable unless a source is linked.
